HAL Id: tel-00872085https://tel.archives-ouvertes.fr/tel-00872085

Submitted on 11 Oct 2013

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Electrodes en diamant pour la fabrication demicrosystèmes électrochimiques pour applications

biologiquesRaphael Kiran

To cite this version:Raphael Kiran. Electrodes en diamant pour la fabrication de microsystèmes électrochimiques pourapplications biologiques. Autre. Université de Grenoble, 2012. Français. <NNT : 2012GRENI077>.<tel-00872085>

Université Joseph Fourier / Université Pierre Mendès France / Université Stendhal / Université de Savoie / Grenoble INP

THÈSE Pour obtenir le grade de

DOCTEUR DE L’UNIVERSITÉ DE GRENOBLE Spécialité: Matériaux, Mécanique, Génie Civil, Electrochimie

Arrêté ministériel: 7 août 2006

Présentée par

«Raphael / KIRAN» Thèse dirigée par «Pascal / Mailley» et «Philippe / Bergonzo» codirigée par «Emmanuel / Scorsone» préparée au sein du Laboratoire Capteurs Diamant au Commissariat à l’Energie Atomique à Saclay

Electrodes en diamant pour la fabrication de microsystèmes électrochimiques pour applications biologiques Thèse soutenue publiquement le «21/09/2012», devant le jury composé de:

Pr. John FOORD Pr. Université d’Oxford, Royaume-Uni, Rapporteur

Pr. Eric McADAMS Pr. Universités INSA, Lyon, France, Rapporteur

Pr. Christoph NEBEL Head of business unit and department, micro- and nano-sensors, IAF

Fraunhofer, Freiburg, Allemagne, Membre

Pr. Etienne GHEERAERT Pr. Université de Joseph Fourier, Grenoble, France, Président

Dr. Pascal Mailley Chercheur au CEA Chambéry, France, Membre

Dr. Philippe BERGONZO Directeur de recherches au CEA Saclay, France, Membre

Dr. Emmanuel SCORSONE Chercheur au CEA Saclay, France, Membre

i

ACKNOWLEDGEMENTS

First and foremost, I would like to affirm that I am immensely indebted to my thesis co-director,

Dr Philippe Bergonzo, director of research at CEA-Saclay, for his inspirational guidance,

constant support and unceasing encouragement for the successful completion of my research

work. Despite his hectic schedule, Philippe has always made himself available whenever I needed

his inputs, comments, and feedback, and helped me to identify and to solve bottlenecks. I feel

extremely privileged to work under him and to learn from his research expertise. It will never

cease to surprise me to think of his profound confidence in my research ability and the magnitude

of freedom, flexibility and opportunity he granted me during all these years of study and analysis

under his guidance. I ever remain infinitely grateful to him.

I gratefully acknowledge my thesis director, Dr Pascal Mailley, formerly from University Joseph

Fourier when my PhD started, and now research scientist at CEA-LITEN in Chambery, for his

invaluable advice and academic support. Pascal has always provided me his esteemed guidance,

starting from my very first days at CEA till the end of my research work, with his profound

knowledge in bio-chemistry, electrochemistry and material science. His vision, caring and

commitment have generously contributed to my formation, work and success.

I have great pleasure in expressing my sincere gratitude to Dr Emmanuel Scorsone who was the

day to day supervisor and who always provided me with the suggestions, questions, and puzzling

remarks I needed to progress. He was with me through all the 3 years and I’m thankful for his

wise supervision and involvement throughout the course of the research.

I am also grateful to Dr Jacques de Sanoit, who introduced me to the fascinating world of

electrochemistry, as well as to several fascinating other worlds where electrolytes close to

Chassagne-Montrachet could be involved. Without his support this work would have never taken

shape. The help, suggestions and constructive criticism received from Prof Jean-Charles

Arnault are also beyond measure and greatly appreciated.

I express sincere thanks to Dr. Lionel Rousseau, research scientist at ESIEE and Prof. Blaise

Yvert , University of Bordeaux for extending their help at various phases of my research and for

providing facilities to carry out part of my research activities at their institutes.

ii

I am indebted to Prof. Richard Jackman, University College London and Prof. Serge Picaud,

INSERM for their valuable suggestions and encouragements throughout my research work.

I would like to extend a very special thanks to my colleagues at CEA-Saclay, Bertrand, Celine,

Christine, Hughes, Michal, Samuel, Alexandre, Amel (Inserm) and to all my friends for their

timely help, co-operation and encouragement during the course of my study. I express my sincere

regret in not mentioning individually many amazing persons who helped me in my pursuit, but I

always nourish appreciation and gratitude for all well-wishers and benefactors.

Last but not least, I would like to extend my special appreciation to my parents for their perennial

and contagious confidence in me and giving me the freedom to choose my career path. I dedicate

my thesis to my beloved parents.

iii

ABSTRACT

Boron doped diamond (BDD) electrodes are extremely promising in the field of biomedical

applications as they exhibit a unique combination of properties. The thesis aims at developing

new types of BDD microelectrodes and exploring their interests for electro-analytical and

electrophysiological applications. Despite their superior electro-analytical properties, BDD

electrodes are prone to fouling, which leads to a loss of electrode reactivity when used in

biological fluids such as urine, waste waters, drinks, blood plasma, etc. A novel electrochemical

treatment was developed to clean the electrode surface and to retrieve the initial reactivity,

thereby enabling the use of BDD electrodes to long periods of measurements without degradation

of the signal, thus significantly extending the field of monitoring and surveying applications up to

domains where continuous analysis is required. The real novelty of the technique is that it does

not require the use of a specific media and thus can be directly performed in the probed (bio-)

fluid.

Microelectrodes in comparison with macro-electrodes offer higher sensitivity, lower background

current, lower ohmic losses and higher signal-to-noise ratio. A robust, high-yield, reliable, and

reproducible process for fabricating a thin-film BDD micro and ultra-microelectrode arrays

(MEA) was developed using a novel lithographic technique, based on clean room processing on 4

inch substrates, thus offering considerable flexibility. For example, among other prototypes,

BDD microelectrodes were developed as biosensors to quantify uric acid in human urine in

quasi-real time. Although diamond film possesses good biocompatibility and excellent

electrochemical properties, the low double-layer capacitance limits its application in

electrophysiological applications. Increasing the charge injection limit was investigated by

surface modification and nano-structuring. These include the synthesis of hybrid diamond-

polypyrrole electrodes and nanograss BDD MEAs. Such high aspect ratio materials appear as

excellent candidates for neurointerfacing applications such as for retinal implants.

iv

v

RESUME

Le diamant dopé bore (BDD) est un matériau extrêmement prometteur pour applications

biomédicales par son unique combinaison de propriétés. Cette thèse a visé le développement de

nouvelles structures de micro-électrodes en BDD et l’étude de leur intérêt et leurs performances

pour des applications électroanalytiques et électrophysiologiques. En dépit de leurs propriétés

électroanalytiques très supérieures à d’autres matériaux d’électrodes plus conventionnels, les

électrodes BDD sont sujettes au «fouling», i.e. l’apparition d’un film à la surface du diamant qui

réduit la réactivité électrochimique. Ceci est très compromettant dans des milieux complexes

comme l’urine, les eaux stagnantes, des boissons, le plasma sanguin etc. Ici, un nouveau

traitement d’activation a été développé pour nettoyer la surface des électrodes et recouvrir leur

réactivité initiale, donc il permet leur usage pour de longues périodes d’enregistrement sans

dégradation du signal. Ceci permet l’usage de ce type d’électrodes, pour des domaines

d’applications, pour le suivi continu d’analytes, sans entretien spécifique, en solutions

complexes. La grande originalité de ces techniques d’activation est qu’elle peut être menée

directement dans l’analyte lui-même.

En comparaison avec leurs équivalents en macro-électrodes, les microélectrodes permettent

d’obtenir de plus grandes sensibilités, des courants résiduels moindre, des pertes ohmiques

moindres, et donc des rapports signal à bruit meilleurs. Un procédé robuste et fiable a été

optimisé pour la fabrication de réseaux de microélectrodes (MEA MicroElectrode Arrays) et

d’ultra micro-électrodes, permettant par lithographie sur 4 pouces d’offrir une large flexibilité de

fabrication. Par exemple, parmi d’autres prototypes, des microélectrodes BDD ont été utilisées

pour applications de biocapteurs pour quantifier l’acide urique en temps quasi-réel. Bien que le

diamant possède une très bonne biocompatibilité et des propriétés électrochimiques excellentes,

la faible relative capacité de double couche limite leur application pour des applications

électrophysiologiques. Des procédés de nanostructuration ont ainsi été mis au point pour accroitre

les limites d’injection de charge. Parmi les approches, des procédés hybrides à base de

polypyrrole se sont révélés prometteurs, de même que des procédés de gravure pour former de la

«nano-herbe» diamant, très intéressantes pour la fabrication de MEAs en BDD. Ces matériaux à

fort rapport d’aspect apparaissent comme d’excellents candidats pour applications d’interfaces

neuronales et notamment pour la fabrication d’implants rétiniens.

vi

vii

TABLE OF CONTENTS

Chapter Page I. INTRODUCTION ..................................................................................................... 1

1.1 Synthesis of diamond ........................................................................................ 3

1.1.1 HPHT synthesis of diamond ....................................................................... 3 1.1.2 CVD synthesis of diamond ......................................................................... 4

1.2 Synthesis of doped diamond ............................................................................. 5 1.3 MPECVD growth at the Diamond Sensors Laboratory .................................... 5 1.4 BDD film characterization ................................................................................ 7

1.4.1 Structural and morphological characterization ........................................... 7 1.4.2 Electrochemical characterization .............................................................. 10

1.5 BDD films for electrochemical sensors and biosensors .................................. 16 1.6 Theory of microelectrodes .............................................................................. 18 Conclusion ............................................................................................................. 21 Bibliography .......................................................................................................... 22 II. Electrochemical activation of diamond electrodes ................................................. 27 Introduction ........................................................................................................... 29 2.1 Ageing and fouling of the electrode ................................................................ 29

2.1.1 Electrode ageing ....................................................................................... 29 2.1.2 Electrode fouling ....................................................................................... 32

2.2 Activation process ........................................................................................... 33 2.3 Activation in other synthetic electrolytes ........................................................ 36 2.4 Influence of pH, current density and number of pulses on activation ............. 38

2.4.1 pH vs activation ........................................................................................ 39 2.4.2 Current density vs activation ..................................................................... 40 2.4.3 Number of pulses vs activation ................................................................. 41

2.5 Surface analysis and activation mechanism .................................................... 42 2.6 In-situ activation in biological fluids .............................................................. 48 Conclusion ............................................................................................................. 50 Bibliography .......................................................................................................... 51 II I. Microelectrode: Design, fabrication and characterization .................................... 53 Introduction ........................................................................................................... 55 3.1 Technological process ..................................................................................... 56

3.1.1 Design 1 .................................................................................................... 56 3.1.2 Design 2 .................................................................................................... 60

3.2 SEM Characterization ..................................................................................... 61 3.3 Electrochemical characterization .................................................................... 63

viii

3.3.1 Background current ................................................................................... 63 3.3.2 Steady state limiting current ..................................................................... 65 3.3.3 Electrochemical impedance spectroscopy ................................................ 67

3.4 All diamond microelectrode arrays ................................................................. 71 Conclusion ............................................................................................................. 75 Bibliography .......................................................................................................... 76 IV. Diamond microelectrodes: Electroanalytical application ..................................... 79 Introduction ........................................................................................................... 81 4.1 Electrochemical characterization of BDD microelectrode .............................. 82 4.2 Cyclic voltammogram of uric acid and ascorbic acid ..................................... 84 4.3 Calibration curves for UA concentration ........................................................ 89

4.3.1 Model 1: Low ascorbic acid concentration ............................................... 89 4.3.2 Model 2: High ascorbic acid concentration .............................................. 91

4.4 Proposed model vs Spectrophotometric quantification ................................... 94 4.5 Automation of quantification procedure and in-situ cleaning ......................... 95 Conclusion ............................................................................................................. 99 Bibliography ........................................................................................................ 100 V. Diamond microelectrodes: Electrophysiological applications ............................. 103 Introduction ......................................................................................................... 105 5.1 Electrophysiological characterization of MEA ............................................. 106

5.1.1 Impedance measurement ......................................................................... 106 5.1.2 Noise level measurement ........................................................................ 108 5.1.3 Neural recording ..................................................................................... 108

5.2 Diamond microelectrode array as neural prosthesis: Retinal implants ......... 110 5.2.1 Platinum microelectrode arrays .............................................................. 111 5.2.2 Diamond microelectrode arrays .............................................................. 114

5.3 Nanograss diamond MEA ............................................................................. 117 5.3.1 Fabrication of nanograss MEA ............................................................... 117 5.3.2 SEM characterization .............................................................................. 117 5.3.3 Electrochemical characterization ............................................................ 118 5.3.4 Electrophysiological characterization ..................................................... 121

Conclusion ........................................................................................................... 122 Bibliography ........................................................................................................ 124 VI. Conclusions and future perspectives ................................................................... 127 Introduction ......................................................................................................... 129 5.1 Electrophysiological characterization of MEA ............................................. 130 APPENDIX A ........................................................................................................... 133

LIST OF TABLES

Table Page Table 1.1 Comparison of silicon and diamond physical characteristics ....................... 3 Table 1.2 Boron doped nanocrystalline layers synthesis conditions ............................. 7 Table 1.3 XPS components for ‘as grown’ BDD film ................................................ 10 Table 2.1 k0 after subsequent scan in human urine ..................................................... 32 Table 2.2 ∆Ep and k0 values of ‘as grown’, aged and activated electrode .................. 35 Table 2.3 k0 after activation in salt solutions .............................................................. 37 Table 2.4 XPS components for ‘as grown’ electrode and after activation .................. 42 Table 2.5 ∆Ep and k0 of electrode with biofilm and after activation ........................... 46 Table 3.1 Mean value and standard deviation of iBG, ilim, and k0 ................................ 69 Table 3.2 Comparison of k0, ilim, CD and iBG of 2 microelectrodes ............................. 71 Table 4.1 UA concentration: Proposed model vs Spectrophotometry ........................ 95 Table 4.2 Comparison of activated and non-activated measurements ........................ 97

LIST OF FIGURES

Figure Page Figure 1.1 Schematics of an MPECVD reactor ............................................................ 6 Figure 1.2 Raman spectra of different BDD films ........................................................ 8 Figure 1.3 Scanning electron microscopic image of BDD film .................................... 8 Figure 1.4 XPS Survey of BNCD thin film .................................................................. 9 Figure 1.5 Schematic diagram of 3 electrodes setup ................................................... 11 Figure 1.6 Schematics of electrical double layer ........................................................ 11 Figure 1.7 Potential window of an ‘as grown’ electrode ............................................ 12 Figure 1.8 CV of an ‘as grown’ BDD electrode in ferri/ferrocyanide ........................ 13 Figure 1.9 Randles equivalent circuit and Nyquist Plot .............................................. 14 Figure 1.10 Nyquist plot of an ‘as grown’ BDD electrode ......................................... 15 Figure 1.11 Schematics illustration of diffusion at a microelectrode ......................... 20 Figure 2.1 CV of an ‘as grown’ and aged electrode in ferri/ferrocyanide .................. 30 Figure 2.2 Nyquist plot of an ‘as grown’ and aged electrode ..................................... 31 Figure 2.3 CV in human urine ..................................................................................... 33 Figure 2.4 Comparison of CV in human urine and after activation ............................ 36 Figure 2.5 Stripping voltammogram for copper, manganese and zinc ....................... 36 Figure 2.6 pH dependence of activation process ......................................................... 39 Figure 2.7 Effect of the current density on the activation process .............................. 40 Figure 2.8 Impact of the number of pulses on the activation process ......................... 41 Figure 2.9 XPS of BDD electrode after activation ...................................................... 43 Figure 2.10 CV of BDD electrode before and after activation ................................... 44 Figure 2.11 Optical microscopic and SEM images biofilm ........................................ 45 Figure 2.12 SEM images of electrode after activation ................................................ 47 Figure 2.13 In-situ activation: k0 after each trial ......................................................... 48 Figure 2.14 In-situ activation: J1 after each trial ........................................................ 49 Figure 3.1 Diamond microelectrode fabrication process (Design 1) .......................... 57 Figure 3.2 SEM images of BDD films on titanium..................................................... 58 Figure 3.3 Potential window and Nyquist plot of BDD films on titanium ................. 59 Figure 3.4 Diamond microelectrode fabrication process (Design 2) .......................... 61 Figure 3.5 SEM image of an 8x8 UMEA along with the tracks ................................. 62 Figure 3.6 Potential window of BDD ultra-microelectrode ........................................ 63 Figure 3.7 RGB model of an 8x8 electrode array: iBG value. ...................................... 64 Figure 3.8 CV of BDD ultra-microelectrode in ferri/ferrocyanide ............................. 65 Figure 3.9 RGB model of an 8x8 electrode array: ilim value ....................................... 66 Figure 3.10 Nyquist plot of BDD ultra-microelectrode .............................................. 67 Figure 3.11 Component of Randles’s circuit of ultra-microelectrode ........................ 68 Figure 3.12 RGB model of an 8x8 electrode array: k0 value ...................................... 69 Figure 3.13 Schematics of interconnected diamond MEA ......................................... 72 Figure 3.14 SEM images of MEA and individual BDD conducting electrode ........... 72 Figure 3.15 CV of boron doped diamond MEA.......................................................... 74

Figure 3.16 CV of boron doped diamond MEA in ferri/ferrocyanide ........................ 74 Figure 4.1 CV and Nyquist plot of BDD microelectrode ........................................... 81 Figure 4.2 CV of uric acid and ascorbic acid in phosphate buffer solution ................ 82 Figure 4.3 Proposed schematics for UA oxidation ..................................................... 83 Figure 4.4 CV of UA at various scan rate ................................................................... 84 Figure 4.5 CV of 1mM UA and AA (0, 250 and 500 µM) ......................................... 86 Figure 4.6 Second order curve fitting: low concentration of AA................................ 87 Figure 4.7 Second order curve fitting: low concentration of AA................................ 88 Figure 4.8 CV of 1.5 mM UA and AA (0, 2 and 4 mM) ............................................ 89 Figure 4.9 Second order curve fitting: high concentration of AA .............................. 90 Figure 4.10 Second order curve fitting: high concentration of AA ............................ 91 Figure 4.11 CV of urine + UA added .......................................................................... 93 Figure 4.12 CV of urine + UA added : with activation ............................................... 94 Figure 4.13 Comparison of activated and non-activated measurements ..................... 95 Figure 5.1 8x8 BDD UMEA fixed on NanoZ device ............................................... 106 Figure 5.2 Magnitude and phase of the impedance: BDD UMEA ........................... 107 Figure 5.3 Mouse embryonic hindbrain-spinal cord on UMEA ............................... 109 Figure 5.4 Recording of neural activity .................................................................... 109 Figure 5.5 Schematics of implantable 8 x 8 Pt microelectrode array ....................... 111 Figure 5.6 Pt soft implant (8 x 8 electrode array) ..................................................... 112 Figure 5.7 CV of implantable Pt microelectrode ...................................................... 112 Figure 5.8 Nyquist plot of implantable Pt microelectrode ........................................ 113 Figure 5.9 CV of implantable Pt in ferri/ferrocyanide .............................................. 113 Figure 5.10 Schematics of implantable 8 x 8 BDD microelectrode array ................ 114 Figure 5.11 CV of implantable BDD microelectrode ............................................... 115 Figure 5.12 Nyquist plot of implantable BDD microelectrode ................................. 116 Figure 5.13 CV of implantable BDD in ferri/ferrocyanide ....................................... 116 Figure 5.14 SEM image of cross-section of BDD nanograss electrode .................... 118 Figure 5.15 Nyquist plot of a nanograss diamond microelectrode ........................... 119 Figure 5.16 CV of nanograss diamond microelectrode ............................................. 120 Figure 5.17 CV of nanograss diamond microelectrode in ferri/ferrocyanide ........... 120 Figure 5.18 Electrode impedance measured for the 4 x 16 array .............................. 121 Figure 5.19 CV of nanograss diamond microelectrode in ferri/ferrocyanide ........... 122

GLOSSARY

µCD: Microcrystalline Diamond

AA: Ascorbic Acid

ARF: Acute Renal Failure

BDD: Boron Doped Diamond

BDD-PPy: Boron doped diamond – Polypyrrole

BNCD: Boron doped NCD

CE: Counter Electrode

CNT: Carbon Nanotube

CV: Cyclic Voltammetry

CVD: Chemical Vapor Deposition

DHAA: Dehydroascorbic Acid

DNP: Diamond Nanoparticles

DPV: Differential pulse Voltammetry

EC: Electrochemical

EDLC: Electric Double Layer Capacitor

EIS: Electrochemical Impedance Spectroscopy

HFCVD: Hot Filament CVD

HPHT: High-pressure/high-temperature

ICU: Intensive Care Unit

MEA: Microelectrode array

MPECVD: Microwave Plasma Enhanced CVD

NCD: Nanocrystalline Diamond

PBS: Phosphate Buffer Saline

PVA: Polyvinyl Alcohol

PVD: Physical Vapor Deposition

RE: Reference Electrode

SEM: Scanning Electron Microscopy

SIMS: Secondary Ion Mass Spectroscopy

SNR: Signal to noise ratio

TBATFB: Tetrabutylammonium tetrafluoroborate

TMB: TriMethyl Borane

UA: Uric Acid

UME: Ultra-microelectrode

UMEA: Ultra-microelectrode array

UNCD: Ultra Nanocrystalline Diamond

WE: Working Electrode

XPS: X-ray Photoelectron Spectroscopy

SYNOPSIS

Electrochemistry is the branch of science which deals with the relationship between electrical and

chemical phenomena and laws of interaction between these phenomena. It is the study of

reactions in which charged particles cross the interface between two phases of matter: electrode

and electrolyte. One can find electrochemical reactions everywhere in nature and even within the

human body. Natural processes such as photosynthesis, respiration, neural transmission are

electrochemical processes. Everyday examples of electrochemistry in our daily life include

batteries, hybrid automobiles, fuel cells, metal extraction, electroplating etc. Electrochemical

sensors have been used extensively as chemical and biomedical sensing units. These

electrochemical sensors include the most commonly used oxygen level monitors, glucose sensors

etc. as well as novel DNA-based sensors. Advances in lithographic and technological processes

have brought revolutionary changes in micro- and nano-structuring of electrochemical sensors.

Recent inventions of novel conducting materials have also contributed towards better electro-

analysis.

Doped diamond is one such material and it has become a major focus of research and

development. Undoped diamond exhibits a high band gap, thus is, at room temperatures,

electrically insulating. When doped with boron, this results in an electrode with remarkable

electro-analytical properties which can be used to detect species in solution before oxygen and

hydrogen evolution interfere with the analysis. Thus, conductive diamond electrodes take

electrochemical detection into new levels and extend their usefulness to analytes which are not

measurable with conventional electrode materials. There are two major approaches to fabricate

doped diamond electrodes: chemical vapor deposition and high pressure high temperature

methods. Finally, diamond, exhibiting specific bio-inert properties, appears as an ideal material

for in-vivo sensing, implantation and other biomedical applications.

This thesis deals with the design, fabrication, characterization and application of boron-doped

diamond electrodes and microelectrodes. Microwave plasma enhanced chemical vapor deposition

is used to fabricate boron-doped nanocrystalline diamond.

This thesis is divided into 6 chapters:

In Chapter 1, the necessary background information and state of the art for this thesis work is

introduced regarding the synthesis of diamond, importance of diamond in electrochemical

sensors and biosensors. In addition, a brief introduction on the importance of micro-structuring

the electrodes is given.

Electrode ageing and fouling are the two major concerns as they can lead to inaccurate readings

and measurement failure. Hence electrode cleaning and activation is necessary. Chapter 2 deals

with novel approaches that were developed during this work which enable the electrochemical

activation of diamond electrodes. A comparison between the existing activation techniques and

our novel pulsed activation technique is described in detail. We have also investigated the

influence of different parameters such as the duration of the pulses, amplitude of current density,

pH of the solution, etc. on electrode reactivity. As the diamond electrodes are targeted for

biomedical applications, an in-situ activation technique was developed where the electrode can be

cleaned within the probed fluid itself. Examples of some in-situ activation techniques in bodily

fluids are also presented.

Microelectrodes possess superior properties over macroelectrodes that are highly favorable for

electro-analysis such as: low ohmic drop, steady-state voltammetric behavior, reduced

background current etc. Chapter 3 is dedicated to the design, fabrication and characterization of

microelectrode and ultra-microelectrode arrays. The chapter describes various design approaches

and the optimization of technological processes in order to produce the microelectrode arrays

with the best performances. The microelectrodes were characterized using scanning electron

microscopy and electrochemical tools including cyclic voltammetry and electrochemical

impedance spectroscopy. Electro-analytical and electrophysiological applications of the

fabricated microelectrodes are described in chapters 4 and 5 respectively.

Chapter 4 is centered on a classical electro-analytical application based on diamond

microelectrode where the uric acid concentration in human urine is selectively determined in the

presence of ascorbic acid. This innovative electrochemical technique is based on fast cyclic

voltammetry and the quantification technique is explained in detail. Chapter 4 also describes the

application of the in-situ activation technique explained in chapter 2, where the electrodes were

reactivated within human urine, thereby opening the way towards automatic quantification of uric

acid quantification with in-situ cleaning.

Chapter 5 focuses on electrophysiological applications of diamond microelectrodes where they

are used to electrically stimulate neurons and to record the neural activity. Although the low

double layer capacitance of a diamond electrode is an advantage in electrochemical recording, the

electrode should possess a large storage capacitance for electrical stimulation. Also for neural

recording, the impedance of the electrode has to be low in order to have higher signal to noise

ratio. The charge injection limit and signal-to-noise ratio were increased by techniques such as

surface modification (boron doped diamond – polypyrrole composite) and nano-structuring

(nanograss). This chapter also describes the fabrication and characterization of flexible diamond

electrodes for eye implants.

The work is summarized in Chapter 6 along with a review of the scope of future studies.

SYNOPSIS EN FRANCAIS

Introduction

Le diamant est un matériau précieux, objet de toutes les convoitises. Sa valeur est conditionnée par sa rareté, sa transparence ainsi que son éclat…Sa présence symbolise bien souvent un amour éternel.

Mise à part son utilisation la plus connue, à savoir la joaillerie, le diamant se trouve également au cœur de l’innovation scientifique. Constitué exclusivement d’un arrangement particulier de carbone, ce matériau est un élément quasiment inaltérable et ses propriétés semi-conductrices en font un matériau recherché en électronique. Il fait également preuve d’une grande résistance à la température et aux rayonnements ionisants. Beaucoup plus isolant que le verre, il acquiert des propriétés conductrices après dopage au le bore (dopage de type p) ainsi il en viendra à posséder une forte réactivité électrochimique avec une large fenêtre de potentiel, pouvant atteindre plus de trois volts en solution aqueuse. Grâce à ce fabuleux Curriculum Vitae, le diamant se trouve positionné au carrefour de multiples technologies innovantes. Le laboratoire LCD du CEA LIST, spécialisé dans la synthèse de ce matériau le décline au sein de plusieurs disciplines comme la dosimétrie médicale, la détection de rayonnements ionisants, la dissipation thermique, les nanoparticules, les transducteurs, les matériaux pour bio interfaces et enfin l’électrochimie analytique.

Dans le cadre de mes études doctorales, j’ai eu l’opportunité d’effectuer mes travaux de recherches au sein du Laboratoire Capteurs Diamant du CEA LIST. L’intérêt particulier que je porte au domaine de la Recherche m’a permis d’appréhender le sujet proposé avec beaucoup de motivation.

Mes travaux sont présentés dans cette thèse, qui est articulée de la manière suivante, après une brève présentation du laboratoire LCD et une introduction au matériau diamant, le chapitre 1 aura d’une part pour but d’éclairer le lecteur sur les méthodes de synthèse du diamant et sur l’importance de ce matériau dans des applications telles que les capteurs électrochimiques et les biocapteurs. D’autre par ce chapitre apportera une brève introduction sur la micro-structuration des électrodes en diamant dopé au bore. Aussi l'un des objectifs de cette thèse a été de développer une méthode de traitement EC (électrochimique) simple et rapide qui peut être utilisée pour récupérer la réactivité perdue des électrode BDD (Diamant dopé au bore), améliorant ainsi leur réutilisation, sans dégradation du signal, même après une longue période de mesures. Le chapitre 2 décrit un nouveau processus d'activation EC, en régime impulsionnel Redox réalisé par une série d’impulsions alternativement cathodiques et anodiques, appliqué sur des électrodes BDD. Ce traitement apportera un changement de la surface de l’électrode, cela peut être mis en évidence par XPS et par caractérisation électrochimique. L’influence de la densité de courant, du pH, et du nombre d'impulsions sur l'activation EC sera développée. Quelques exemples d'activation dans les fluides biologiques et synthétiques seront également inclus dans

ce chapitre. Ensuite le chapitre 3 sera consacré à l’explication des différents processus de fabrication des microélectrodes et des MEA. Ces électrodes ont été caractérisées par MEB et par des méthodes électrochimiques (Voltamétrie cyclique CV et spectroscopie impédance électrochimique EIS). Cela à permis d’apprécier les performances des électrodes à savoir les courants limites (ilim), les constantes de transfert de charge (k0), les fenêtres électrochimiques et les courants de fond (iBG). Ces deux méthodes d’analyses électrochimiques sont les techniques les plus efficaces pour détecter d'éventuelles fissures ou discontinuités dans la couche isolante de l’électrode BDD UME, ainsi que pour l'analyse de ses propriétés électrochimiques. En ce qui concerne le chapitre 4, il traitera les études effectuées sur la microélectrode BDD. La microélectrode BDD d’un diamètre de 40 µm et en forme de bande, fabriquée selon notre technologie (voir chapitre 4), a été utilisée pour la détection sélective et sensitive de l'acide urique (UA) en présence de quantités faibles ou élevées d’acide ascorbique (AA). La capacité de la double couche de diamant, réduit le courant de fond et augmente le signal vers le ratio initial. Ces microélectrodes indiquent une diminution de la chute ohmique, une couche de diffusion hémisphérique et permettent la mise en place rapide d'un signal statique en comparaison avec les macro-électrodes. Ce chapitre traite de la caractérisation électrochimique de ces microélectrodes, de la quantification de l'acide urique en présence d'une grande ou faible concentration d’acide ascorbique et de différentes techniques de nettoyage in-situ. Enfin l’électrophysiologie est l'étude des propriétés électriques de cellules et de tissus biologiques, elle implique la mesure de la variation de tension d'une entité biologique. Il s'agit d'une approche puissante qui permet d’une part l’étude de l'activité électrique des cellules animales, pour comprendre le fonctionnement du système nerveux, du cerveau ou de l'hypothalamus, et d’autre part le diagnostique et le traitement des troubles du système nerveux. Des réseaux de microélectrodes extracellulaires permettent de mesurer directement l’activité électrique des neurones. Les MEA fabriquées en utilisant la technique décrite dans le chapitre 3, ont été utilisées pour des mesures électro-physiologiques in-vitro. Le chapitre 5 décrit le processus de fabrication et la méthode de caractérisation d'une prothèse neurale: implants rétiniens constitués de MEA BDD. Les études sur l'amélioration de la limite d'injection de charges sont aussi incluses dans ce chapitre, elles se font par modification de surface et nano-structuration de l'électrode.

Le Laboratoire Capteurs Diamant du CEA

Le CEA :

Le CEA est un organisme public de recherche créé en 1945 par une ordonnance du général de Gaulle. Cette institution est actuellement un acteur majeur dans la recherche française et européenne et pionnier en matière d’innovations technologiques. Elle emploie actuellement plus de 15 000 salariés sur ses centres répartis dans toute la France.

Son domaine d’activité se situe dans cinq grands domaines qui sont :

- le nucléaire civil (développement des technologies de réacteurs, retraitement des combustibles nucléaires usés et traitement des déchets nucléaires)

- la recherche technologique (développement des technologies du futur en partenariat avec des groupes industriels : cela concerne le développement de nouveaux matériaux, les technologies pour l’information, la communication, la santé…)

- la défense nationale (armes atomiques, systèmes de propulsion des navires et sous-marins à propulsion nucléaire)

- sciences de la matière et de l’univers (recherches fondamentales sur la structure de l’univers, les nanosciences, l’énergie de fusion, l’astrophysique…)

- sciences du vivant (amélioration de l’imagerie médicale, biologie moléculaire…)

Le laboratoire Capteur Diamant (LCD) : Le laboratoire LCD est axé sur la recherche technologique pour les applications industrielles et biomédicales. Il est rattaché à un département nommé DCSI (Département Capteurs, Signal et Information) lui-même étant un élément du LIST (laboratoire intégration des systèmes et des technologies). La particularité du LCD concerne l’étude du diamant de synthèse pour le développement de capteurs innovants (chimique, rayonnement, biologique, radioactivité…). Actuellement, vingt-sept personnes travaillent au LCD ; quatorze permanents, le reste de l’équipe est constitué de post doc, thésards, ingénieurs en alternance et stagiaires. Le matériau diamant

Structure du diamant

Le diamant est l’une des trois formes allotropiques du carbone avec le graphite et le fullerène. Alors que le graphite se trouve sous forme de feuillets hexagonaux et le fullerène sous la forme de sphères, d’un ellipsoïde ou d’un tube composé de feuillets, le diamant se présente sous la structure cristalline cubique faces centrées (CFC). La structure diamant comporte huit atomes de carbone par maille élémentaire.

Chaque atome de carbone est lié de façon covalente aux quatre autres: ils sont dits hybridés sp3. La forme sp3 est une forme métastable du carbone; la forme stable du carbone étant le graphite où le carbone trivalent a une hybridation sp2.

Figure 1 : Maille élémentaire du diamant

A l’état naturel, le diamant se forme sous la croûte terrestre à des profondeurs telles que les températures et pressions atteignent respectivement 1100 à 1400°C sous 4 à 6 GPa, conditions nécessaires à la cristallisation en une structure dérivée du CFC du carbone qui le compose.

Propriétés du diamant

Le réseau cristallin du diamant est très dense (1,54 Å entre chaque atome voisin). Ce matériau possède un fort nombre d’atomes par unité de volume (1,76 1023 atomes.cm-3). Les fortes énergies de cohésion atomiques (7,3 eV/atome) et sa très faible distance interatomique font du diamant un matériau très résistant, électriquement isolant et très bon conducteur thermique (20 W.cm-1.K-1, 5 fois plus élevée que celle du cuivre). Il est inerte vis-à-vis d’agents agressifs aussi divers que sont les acides, les bases, les oxydants et réducteurs. Le diamant est également très résistant aux rayonnements ionisants (dose intégrée maximale de 2,5 MGy pour le rayonnement gamma de 1 MeV). De plus, son utilisation dans le domaine de la dosimétrie médicale est envisageable car son numéro atomique (Z = 6) est très proche de celui du tissu humain (Z = 7,1) [Tromson, 2000]

.

Le tableau ci-dessous récapitule les principales caractéristiques physiques du diamant et les compare avec celles du silicium [Jany, 1998].

Caractéristiques Diamant Silicium

Durée de vie des porteurs de charges

100 ps à 10 ns pour le diamant polycristallin et 30 ns pour le

diamant monocristallin 0,10 s

Densité atomique 1,76.1023 atomes.cm-3 4,96.1022

atomes.cm-3

Densité 3,51 2,33

Largeur de bande interdite

5,5 eV 1,1 eV

Résistivité de 1012 .cm à 1015 .cm 5.105 .cm Mobilité des électrons à 300 K

2000 cm2.V-1.s-1 1350 cm2.V-1.s-1

Énergie de cohésion 7,37 eV par atome 4,63 eV par atome

Énergie de création paire électron/trou

13 eV 3,6 eV

Température maximale d’utilisation

> 500 °C 50 °C

Tableau 1

Caractéristiques physiques du diamant comparées à celles du silicium

Le diamant naturel est classifié en fonction de sa teneur en impuretés:

Type Ia : 0,1 % d'azote sous forme d'agrégats. La majeure partie (98 %) des diamants extraits des mines sont de ce type.

Type Ib : azote en position substitutionnelle, couleur jaune.

Type IIa : très faible quantité d'azote. Il représente moins de 1 % de la production mondiale.

Type IIb : diamant semi-conducteur. La conduction est due à la présence de bore, qui donne une couleur bleutée au diamant. Il est extrêmement rare dans la nature (< 0,1 %).

La fabrication de capteurs à partir de diamants naturels est limitée par différents facteurs dont l'impossibilité de trouver deux diamants naturels de caractéristiques physiques identiques (répartition et concentration différentes en impuretés et défauts).

Afin de contrôler les impuretés et la taille du diamant, différentes techniques de synthèses ont pu être développées. Au laboratoire LCD, la technique MPCVD (Dépôt chimique en phase vapeur assistée par plasma micro-onde) est utilisée pour la synthèse du diamant dopé au bore (Dopage type p).

Le diamant poly-cristallin est synthétisé sur un substrat en silicium, ce qui permet de réaliser des films de grande surface à moindre coût (jusqu’à 10 cm de diamètre). La structure microscopique du diamant poly-cristallin est composée de grains.

Contexte de l’étude

Le diamant de synthèse est très prisé dans le monde de la recherche. Des techniques de synthèse ont été perfectionnées lors de ces quarante dernières années. Ces avancées scientifiques ont rendu ce matériau disponible [Bundy, 1955].

Dans le cadre de cette étude, nous nous intéressons au diamant de synthèse (procédé CVD assisté par micro-ondes) et plus précisément celui dopé au bore électrochimiquement actif. En effet, le diamant intrinsèque est un matériau semi-conducteur avec un grand gap (5.5 V) et pour qu’il puisse être utilisé en électrochimie, un dopage est nécessaire afin de le rendre conducteur électrique. Grâce à ce dopage (dopage au bore de type p), ce matériau devient électro-actif, une nouvelle propriété qui figure dans une longue liste d’autres caractéristiques à la fois exceptionnelles et originales, notamment une large fenêtre de potentiel (plus de 3 V en milieu aqueux), de bas courants résiduels (quelques µA), une surtension importante à la réduction de l’oxygène dissout ainsi qu’une impressionnante stabilité chimique [Fujishima, 2005, Kraft, 2007, Pleskov, 2006]. Plus récemment c’est pour sa bio-inertie que le diamant se voit propulsé dans la sphère de la communauté scientifique. Cette propriété élargit encore le vaste champ des applications du diamant qui s’étend de la détection de substances explosives et polluantes à l’état de traces [de Sanoit, 2009] à la réalisation de prothèses neuronales [Artifical Rectina, 2001, DREAMS, 2006, MEDINAS, 2008]. Cependant, pour pouvoir réaliser toutes ces applications, l’expérimentateur doit posséder une bonne connaissance et une maitrise parfaite de la synthèse du diamant et de son dopage. C’est dans ce cadre qu’il est proposé d’étudier les propriétés du diamant avec différents taux de dopage au bore et de les caractériser par les méthodes citées plus bas.

Synthèse du diamant

Au vu des propriétés intéressantes du diamant pour la recherche et l’industrie, les chercheurs se sont naturellement intéressés à sa synthèse. Le premier diamant artificiel provient de Stockholm et date de 1953 [E. Vanhove, 2010].

Dans le cadre de cette étude la synthèse du diamant se fera par dépôt en phase vapeur assisté par plasma micro-onde (MPCVD) dans une enceinte métallique sous vide qui constitue le réacteur. Le réacteur est identifié au laboratoire sous le nom de BAOBAB .

La méthode MPCVD nécessite au moins deux gaz, le premier apporte les atomes de carbone, le second les atomes d’hydrogène. Usuellement pour l’apport de carbone, le méthane (CH4) est utilisé, quant à l’hydrogène (H2) il sera fourni par le

second gaz (dihydrogène H2). La pression de l’enceinte est comprise entre 30 et 100 mbar.

Ces deux gaz forment un mélange gazeux qui sera ionisé grâce à un champ électrique intense. Ce champ électrique est produit par l’énergie micro-onde. Le mélange gazeux se transforme alors en plasma. Plusieurs facteurs influencent la qualité du matériau produit : la pression dans l’enceinte, la température du substrat (900 °C), la puissance micro-onde (quelque kW) et la composition du mélange gazeux. (Cf figure 2)

Figure.2 : Schéma de principe de la méthode MPCVD

La première étape de synthèse est la nucléation. Elle consiste à créer des germes de diamant sur un substrat de silicium. Ce dernier est utilisé sous la forme de disques de 5 cm de diamètre recouvert de nanoparticules de diamant par spin coating afin d’initier la nucléation.

A partir des sites de nucléation, les germes de diamant croissent pour finalement se rejoindre (coalescence) et ainsi former un dépôt poly-cristallin (Cf. figure 3).

Une nouvelle phase débute alors, elle est nommée phase de croissance du film. Les atomes de carbone vont venir se lier à la surface des cristaux afin de les faire grandir.

Le film possède une structure colonnaire (Cf. figure 3).

Gaz : CH4, H2

Couche de

diamant Substrat

de Silicium

Plasma

Micro-ondes

Figure. 3: Coalescence des cristaux de diamant et structure colonnaire

obtenue après croissance du film

Le but de cette étude étant l’utilisation du diamant de synthèse pour des applications de type capteurs électrochimiques, il est nécessaire de le doper au bore afin qu’il acquière les propriétés de conduction électrique et de réactivité électrochimique. Pour se faire, il sera dopé au bore (dopage de type p) via l’injection du gaz TMB (triméthylbore) lors de la croissance. La concentration en bore qui sera inclus dans le matériau sera contrôlée par l’addition d’oxygène sous forme gazeuse (avec un gradient d’O2/H2 de 0 à 0.5 %) à débit de TMB constant. (Cf. le tableau 2, pages 10)

Caractérisation du matériau diamant dopé au bore

Nous avons choisi deux méthodes pour caractériser la morphologie du diamant dopé au bore. La première est la microscopie électronique à balayage (MEB) qui permet l’observation de l’évolution de la structure poly-cristalline pour différentes conditions de croissance. La deuxième méthode choisie est la spectrométrie de masse d’ions secondaires (SIMS) qui a pour but de mesurer le taux de dopage en bore dans l’épaisseur du matériau. (cf. annexe 4°).

Caractérisation électrochimique

La caractérisation électrochimique a pour but de tester la réactivité de surface du diamant de synthèse dopé au bore afin d’en déterminer la propriété fondamentale (vitesse de transfert de charge k0 mesurée avec un couple redox rapide). Deux méthodes ont été choisies; une méthode dite « à balayage » nommée voltamètrie cyclique et une méthode dite « à modulation » nommée spectroscopie d’impédance électrochimique .

Conditions de croissance du diamant dans le réacteur Baobab type MPCVD

Dans le tableau ci-dessous, sont regroupés, les conditions de croissance des échantillons de diamant de synthèse sur substrats de 5 cm de diamètre, l’épaisseur de la couche diamant obtenue, ainsi que, la valeur du taux dopage en bore mesuré par SIMS. Le but recherché est de vérifier qu’il est possible de contrôler le taux de dopage en bore du diamant en faisant varier la concentration en oxygène dans le réacteur à concentration de TMB constante.

Dépôts H2/CH4

(sccm)

TMB

(sccm)

O2 (sccm)

Puissance/Pression

Température (°C)

Epaisseur

(nm) pesée

Dopage

estimé/

SIMS B1203

12 100 /

1 12 0 30 mbar

1.5 kw 800 280 2.10

20

B150312-a

100 / 1

12 0.15 33 mbar 1.5 kwatts

820 320 4.10

19

B130312-a

100 / 1

12 0.25 30 mbar 1.5 kw

800 290 2.1019

B150312-b

100 / 1

12 0.35 33 mbar 1.5 kw

790 370 8.1018

B130312-b

100 / 1

12 0.5 35 mbars 1.65 kw

810 330 5.1018

B240212

100 / 1

12 1 35 mbar 1.65 kwatts

750 350 2.1017

Tableau 2 Conditions de croissance des échantillons de diamant utilisés lors des expériences.

.

Mise en œuvre d’une électrode de diamant

Les films de diamant sortis du réacteur Baobab après caractérisation par SIMS et MEB sont montés sous la forme d’électrodes afin de pouvoir être caractérisés par les méthodes électrochimiques CV et EIS. Le montage des électrodes requiert un certain nombre d’éléments (Cf. Figure 4). Les différentes étapes du montage des électrodes sont décrites dans le mode opératoire suivant :

Figure 4 : Schéma éclaté de l’électrode de diamant de synthèse dopé au bore montée

Etape 1 : Prendre une lame de microscope en verre et la couper en deux parties dans le sens longitudinal. Nettoyer un des morceaux à l’éthanol absolu (95 %) afin d’enlever toutes traces de graisse. Coller une bande de cuivre adhésive sur une des faces afin de former un contact électrique. Eliminer le surplus de bande cuivre afin d’obtenir un assemblage de même type que celui montré ci-dessus. Effectuer un dernier nettoyage à l’éthanol.

Etape 2 : Prendre à la manière d’une plume d’écriture un peu de mélange eutectique indium/gallium à l’aide d’un trombone déplié. Gratter vigoureusement la surface du cuivre sur environ 1 cm2 afin de former un alliage avec ce dernier. Pratiquer de la même façon sur la surface arrière (opposée à celle du diamant) du substrat de silicium afin d’éliminer la couche superficielle de SiO2 qui sera alors remplacée par un fin dépôt d’alliage eutectique. Cette technique particulière permet d’assurer un contact de type ohmique dans l’électrode.

Etape 3 : A l’aide d’une pince brucelles, poser délicatement le substrat (coté silice et eutectique) sur le cuivre recouvert d’eutectique indium/gallium. A l’aide d’un coton tige, on appuiera délicatement sur la surface de diamant afin de solidariser l’ensemble du montage.

Etape 4 : Enrober le montage à l’aide d’une résine époxy (Araldite rapide bi-composants) tout en ménageant une fenêtre de diamant et en laissant à nu le haut du contact électrique en cuivre. Laisser polymériser 24 heures à température ambiante et à l’abri de la poussière

Araldite

Diamant

Silicium

In/Ga

Verre

Cuivre

Araldite

Diamant

Silicium

In/Ga

Verre

Cuivre

Mesure de la surface de l’électrode

La surface des électrodes a été déterminée par une photographie des électrodes posées sur une mire en papier quadrillé. Après découpage des images obtenues et d’une surface de référence, nous avons pesé les différents échantillons de papier sur une balance de précision. La mesure de masse des photographies des électrodes et des échantillons de référence de surface connue a permis de déterminer l’aire des électrodes de diamant. Nous avons admis que le grammage du papier était constant

Caractérisations électrochimiques

Les caractérisations électrochimiques ont été réalisées d’une part par spectroscopie d’impédance électrochimique (EIS) au potentiel de repos et d’autre part par voltamètrie cyclique. L’électrolyte utilisé est composé dans tous les cas par une solution équimolaire de ferro-ferricyanure de potassium (K3Fe(CN)6 et K4Fe(CN)6) à la concentration de 10-3 M. Le sel de fond employé est le chlorure de potassium (KCl) à la concentration de 0.5 M.

L’activation électrochimique des électrodes de diamant dopées au bore a été effectuée en régime impulsionnel (± 5 mA, 0.1 s, 30 cycles) dans un électrolyte composé de Coca Cola Light Décaféiné.

Protocole opératoire

Le protocole opératoire employé pour l’étude des différents échantillons de diamant dopé au bore a été le suivant :

1) Mesure EIS du matériau « as grown » afin de mesurer la réactivité initiale du matériau. Trois mesures successives sont effectuées afin d’évaluer la reproductibilité ou de mettre en évidence une évolution de la réactivité de l’électrode en cours de mesure.

2) Mesure par voltamètrie cyclique entre - 0.3 et + 0.3 V/ECS dans une gamme de vitesses de balayage s’étendant de 25 à 150 mV/s. Pour une vitesse de balayage donnée, quatre cycles de mesure sont effectués. Seul le dernier cycle sera systématiquement enregistré.

3) Mesure EIS « Après CV » afin de comparer la réponse de l’électrode avant et après la mesure de voltamètrie cyclique. Trois mesures successives sont effectuées afin d’évaluer la reproductibilité de la réponse de l’électrode.

4) Activation électrochimique du matériau en régime impulsionnel Redox par

une série d’impulsions alternativement cathodiques et anodiques (± 5 mA, 0.1 s, 30 cycles) dans un électrolyte composé de Coca Cola Light Décaféiné.

5) Mesure EIS « Après activation électrochimique » afin de comparer la réponse de l’électrode avant et après l’activation électrochimique. Trois mesures successives sont effectuées afin d’évaluer la reproductibilité de la réponse de l’électrode.

6) Mesure de voltamètrie cyclique entre - 0.3 et + 0.3 V/ECS dans une gamme de vitesses de balayage s’étendant de 25 à 150 mV/s. Ceci permettra de comparer la réponse de l’électrode en voltamètrie cyclique « avant » et « après activation électrochimique ». Pour une vitesse de balayage donnée, quatre cycles de mesure sont effectués. Seul le dernier cycle sera systématiquement enregistré.

7) Mesure EIS « Après activation électrochimique et CV » afin de comparer la réponse de l’électrode « après l’activation électrochimique » et « après activation + CV ». Trois mesures successives sont effectuées afin d’évaluer la reproductibilité de la réponse de l’électrode.

L’activation électrochimique

L’activation électrochimique ne semble pas efficace pour des teneurs en bore dans le matériau minimal. En effet, en dessous d’un taux de dopage de 1020 B/cm3 la méthode devient inefficace voir nuisible pour l’intégrité des propriétés électrochimiques des électrodes.

0,0 0,2 0,4 0,6 0,8 1,0

-4

0

4

8

E (

V/A

g-A

gC

l)

Time (s)

B 300112

Coca Cola Light décaféiné

Pulses de 5 mA pendant 0,1s

30 cycles

Figure 5 Un exemple d’activation électrochimique impulsionnelle

d’une électrode de diamant dopée au bore :

La constante de vitesse de transfert de charges k0

La réactivité électrochimique d’un matériau électro-actif est estimée grâce à l’indicateur constante de vitesse de transfert de charges, noté k0 et de dimension cm-1. Cet indicateur, intégrant dans son calcul à la fois la résistance de transfert de charges et la surface active des électrodes, il sera aisé de comparer les propriétés électrochimiques d’électrodes de surfaces différentes. On estime que la réactivité d’une électrode mesurée à l’aide d’un couple électrochimique rapide (i.e Fe(CN)6

3-/4-) est satisfaisante si la valeur de k0 se trouve être supérieure à 10-3 cm/s.

Conclusion

Dans le premier chapitre, les différentes techniques de croissance du diamant dopé au bore et les méthodes de caractérisation de ce matériau ont été décrites. Les électrodes de diamant dopé au bore ont été synthétisées à partir d'un mélange gazeux comprenant du méthane, de l’hydrogène et du TMB (tri-méthyl borane) par méthode MPCVD (dépôt en phase vapeur assisté par plasma micro-onde). Les propriétés physiques et chimiques de la couche mince du diamant synthétisé, ont été évaluées en utilisant la spectroscopie Raman, XPS, MEB, CV et EIS. Les électrodes de diamant fortement dopé au bore présentent une excellente réactivité, ce qui en fait des candidats idéaux pour des applications électro-analytiques et des applications en tant que biocapteurs.

Il a été démontré, que la méthode d’activation développer pendant mes recherches, permet de récupérer la réactivité de l’électrode BBD perdue au court du temps. Cette perte de réactivité est due au contact de l’électrode avec l’air ou encore à un encrassement de cette dernière par une solution. En ce qui concerne cette méthode d’activation impulsionnelle, plus la densité de courant sera élevée et plus le temps d’activation sera court, meilleur sera le résultat. En ajustant les paramètres mentionnés ci-dessus (densité de courant, durée d'impulsion, nombre d'impulsions et type d’électrolyte), on peut augmenter la constance de transfert de charge k0, pour atteindre des valeurs supérieures à 0,01 cm.s-1. L'autre avantage de cette technique est qu’on peut réutiliser l'électrode BDD sans craindre de perdre ses caractéristiques. Contrairement à d'autres prétraitements plus classiques rapportés dans la littérature, tels que les traitements anodiques, cathodiques ou thermiques, ce nouveau prétraitement électrochimique est relativement simple, rapide et nécessite un minimum d'énergie.

Un nouveau procédé lithographique, reproductible et avec un rendement élevé, a été utilisé pour fabriquer les UMEA. Des BDD UMEA, appropriées pour une utilisation dans des capteurs électrochimiques, ont été préparées par une technique de micro-fabrication compatible avec la technologie standard de salle blanche. La caractérisation topographique détaillée et l’étude électrochimique des UME individuelles n’ont révélé que quelques électrodes défectueuses, dans une plateforme contenant plusieurs microélectrodes. Pendant les tests électrochimiques, les UME ont été exposées à un faible courant de base, presque théorique, à un état d'équilibre limitant et à un taux de transfert d'électrons rapide (aux alentours de 0,01 cm s-1). L'amélioration de ces deux valeurs physiques a été réalisée grâce à l'activation électrochimique. L'objectif de ce travail, a été de développer des plateformes de bio-détection pour le suivi d’activités neuronales en l'électrophysiologie. Les applications électro-analytiques et électro-physiologiques de ces microélectrodes et de ces plateformes AME seront plus détaillées dans les chapitres 4 et 5.

La détermination sélective de l’acide urique en présence d'acide ascorbique a été réalisée en utilisant la microélectrode BDD sans y apporter des modifications. La comparaison de la technique de quantification électrochimique EC et de la technique

spectrophotométrique, montre qu'une mesure précise peut être faite en utilisant les deux modèles proposés. Cette technique met en évidence que le potentiel, des microélectrodes BDD considérée comme un capteur électro-analytique, est du à la faible capacité de la double couche de l’électrode qui résiste à une haute densité de courant et qui possède une résistance à la corrosion. Le traitement électrochimique récupère la réactivité perdue d'une électrode qui a été soit séchée à l'air, soit encrassée par un milieu avec lequel elle aurai été mise en contacte. Pour ce traitement, il n’est pas utile d’utiliser un réactif ou une solution spécifique. Cette technique permet la réutilisation de la microélectrode BDD et même son activation dans l'urine même. Cela démontre la possibilité d'automatisation de la quantification de l’acide urique car l’électrode peut être activé directement dans l'urine, donc elle peut être utilisée pour une surveillance continue et pendant une longue période d’analyse. Le temps nécessaire pour l'activation est de 300 ms et le temps nécessaire pour la voltamétrie cyclique CV (20 Vs-1) est inférieur à 200 ms.

Le diamant peut être une prothèse neurale avec d’excellentes propriétés physiques et chimiques et possède aussi une bonne biocompatibilité. Dans ce chapitre, plusieurs approches ont été proposées pour améliorer les propriétés électriques du diamant afin d’en faire une électrode idéale pour des applications électro-physiologiques. Bien que les microélectrodes BDD-ppy électrodes ne peuvent pas être utilisés en tant que prothèse neurale, cette étude a permis l’utilisation de ce matériau dans d’autres domaines de recherche tels que le stockage d'énergie et de transmission. Les microélectrodes BDD nanograss, présentent de très larges limites d'injection de charge et par rapport au nitrure de titane et l'oxyde d'iridium, le diamant possède une meilleure biocompatibilité et stabilité microstructurale. L'augmentation du nombre d'électrodes, la conception des formes d'électrodes plus appropriées (jusqu'à présent, les formes étaient toutes planaires) et une modification surface de l'électrode peuvent faire des microélectrodes BDD nanograss un bien meilleur dispositif avec des propriétés supérieures à d’autres techniques usuellement utilisées pour la fabrication de prothèses neuronales.

Les résultats des études présentées dans cette thèse suggèrent des intérêts au cœur de l’innovation scientifique et technologique. En ce qui concerne le processus d'activation in situ décrit dans cette thèse, il peut trouver des applications dans de nombreux processus d'analyses on peut citer; la détection de neurotransmetteurs comme la dopamine ou encore la catécholamine, quant aux analyses in-vitro, plusieurs applications sont possibles car les microélectrodes BDD sont connues pour leur bio-inertie. D'autres applications incluent l'activation in situ, cette méthode permet de quantifier la teneur totale en polyphénols au cours de la fermentation du vin ou encore le traitement des eaux usées. Les microélectrodes BDD en bande et ceux en réseau MEA peuvent être utilisées pour d'innombrables applications biomédicales telles que la détection de métaux lourds, de maladies neurochimiques ou encore pour des enregistrement électriques qui trouvent leur utilité dans l’étude de l’activité neuronale, aussi ces microélectrodes peuvent servir à l'administration de médicaments, à la fabrication de prothèses pour des implants rétiniens et cochléaires.

Enfin la fonctionnalisation de ces microélectrodes par des enzymes pourra permettre d’étendre le rayon de ces applications en tant que capteur de glucose, d’alcool ou d’ADN.

CHAPTER I

Introduction

Introduction

Chapter I Page 2

Introduction

Chapter I Page 3

Diamond is an allotrope of carbon where the carbon atoms are arranged in a strong tetrahedral

structure. The three dimensional assembly of different sp3 hybridized carbon atoms, where the

bonds are along the four directions connecting the center of a regular tetrahedron to its four

corners, makes diamond the hardest natural material known. Naturally occurring diamond

possesses excellent properties such as high thermal conductivity, optical transparency,

electrical insulation, chemical corrosiveness, and biocompatibility.1 Intrinsic diamond is a

wide band gap material with Eg = 5.45 eV.2 Table 1.1 compares the physical properties of

silicon, a classical semiconductor, with those of diamond, a wide band gap semiconductor.

Diamond’s excellent properties make it an ideal candidate for electronic devices.

Table 1.1 Comparison of silicon and diamond physical characteristics2

Property Si Diamond

Band gap (eV) 1.12 5.45

Dielectric constant 11.9 5.5

Electric breakdown field (kV.cm-1) 300 10000

Electron mobility (cm².V-1.s-1) 1500 2200

Hole mobility (cm².V-1.s-1) 600 850

Thermal conductivity (W.cm-1.K-1) 1.5 22

1.1 Synthesis of diamond

Diamond can be synthesized using two techniques: High-pressure/high-temperature (HPHT)

and chemical vapor deposition (CVD). Large single crystal diamonds can be grown by HPHT

techniques whereas CVD techniques produce mainly polycrystalline diamond.3 The

description of these two growth techniques are described in sections 1.1.1 and 1.1.2.

1.1.1 HPHT synthesis of diamond

In nature, diamonds are formed at high temperature and high pressure deep in the Earth

mantle. The high-pressure high-temperature (HPHT) process essentially simulates nature's

process for synthesizing diamonds. The HPHT method uses a molten metal catalyst (usually

nickel, cobalt or iron) to facilitate the change from graphite to diamond. A reproducible

diamond synthesis technique was first demonstrated in 1955 at the General Electric

Company.4 In their process, diamond was synthesized using a high pressure of 75000 atm,

operating temperatures from 1200 to 2000 °C and heating pulses, generated by discharging a

Introduction

Chapter I Page 4

large electrolytic capacitor through the graphite sample (in order to avoid the melting or

chemical reaction of the surrounding wall material). Although non-metallic catalysts like

oxides, hydroxides and halides can also be used to synthesize diamond, however, such

process requires higher pressure-temperature budgets, thus longer reaction times are required.

The HPHT method is currently the most commonly used technique to fabricate industrial

diamond (polishing, cutting tools, etc.). They are also used to change or enhance the colors of

some rare natural diamonds by this technique, thus making them more valuable in the market.

1.1.2 CVD synthesis of diamond

The logic behind CVD processes lies in the thermal decomposition of carbon containing gases

on natural diamond crystals heated between 600 and 1600 °C.4 At higher temperature

excessive black carbon is deposited and at lower temperature insufficient growth takes place

and hence the temperature range should be 900 – 1100 °C. Essentially, the gas phase species

are energetically activated by either hot filament (HFCVD) or microwave plasma

(MPECVD).5 Eversole was the first to demonstrate CVD diamond growth at low pressure.6

Matsumoto et al. made a significant breakthrough by demonstrating that atomic hydrogen

etches sp2 graphitic phases.7 In addition to the aforementioned activation techniques, DC-

plasma, RF-plasma, electron cyclotron resonance-microwave plasma CVD (ECR-MPCVD),

and their modifications were developed.6,8

The critical step in CVD diamond growth is the nucleation phase. Diamond single crystals

were used as the substrate to grow diamond during the early development of CVD diamond

deposition. Later diamond deposition based on diamond seeds and hetero-substrates were

developed. Mitsuda et al. found that scratching of the substrate surface with diamond powder

could greatly enhance the nucleation density.9 Yugo et al. reported on generation of the

diamond nuclei from ion bombardment in an electric field in plasma chemical vapor

deposition.10 Methane (CH4) is the commonly used precursor which is blended with a large

quantity of hydrogen (H2). Many active species (CH4+, CH3

+, CH2+, CH+, C+, H+, H2

+, CH5+

and H3+) are produced and they react with the surface of the diamond crystals.11 Both

amorphous and diamond allotropic forms of carbon are grown together during deposition. The

monoatomic hydrogen species play an important role in etching the amorphous phase. Based

on the grain size, the polycrystalline diamond films can be classified as microcrystalline (µCD

- > 1 µm), nanocrystalline (NCD – few 100 nm) and ultra-nanocrystalline (UNCD - < 100

nm) diamond.6,12,13

Introduction

Chapter I Page 5

1.2 Synthesis of doped diamond

Although diamond is one of the best natural insulators, when doped, the material possess

metallic behavior, thereby making it an excellent material for electrochemical, electronic and

optoelectronic devices. Doping of diamond is achieved by introducing dopant atoms into the

plasma during growth. The growth of boron doped diamond (BDD) films by CVD techniques

has been achieved using a boron source added to the gaseous mixture resulting in p-type

diamond films.14 N-type doping of diamond has been achieved by several techniques

(nitrogen atom incorporation,12 diphosphorous pentoside15 lithium ion implantation,16 etc.)

which are more complicated as opposed to p-type doping. Koizumi et al. has reported on n-

type semiconducting monocrystalline diamond using phosphine as the dopant source.16

1.3 MPECVD growth at the Diamond Sensors Laboratory

The Diamond sensors Laboratory (LCD) of CEA-LIST houses several dedicated MPECVD

reactors enabling the growth of intrinsic and doped diamond layers (on substrates up to 4” in

diameter) as well as heteroepitaxial growth. Two reactors were particularly used to grow the

BDD films used during this PhD thesis:

1) Baobab: A home-made, cylindrical-shaped metal reactor exhibiting a plasma

configuration close to that of the ASTeX PDS type reactors.

2) Seki: ASTeX AX6500 series reactor

Although several substrates such as quartz and highly resistive silicon wafers were used to

grow BDD films, p-type silicon wafers (Siltronix) were employed for most of our

experiments. They are monocrystalline, polished, ‘100’ oriented and heavily boron-doped

(resistivity below 0.05 ohm.cm) silicon wafers with a thickness of 500 microns. The silicon

substrates were sonicated in isopropanol and were then rinsed thoroughly with ultra-pure

deionized water (18 Mohm) and dried under an argon stream. Diamond nucleation was

initiated by diamond particles deposited on the surface of the substrate using nano-seeding.

Aqueous, colloid-containing, diamond nanoparticles (DNP) purchased from Van Moppes

were spread by spin coating to achieve a high density nanodiamond seeding using a method

developed previously in the lab (refer to chapter 3 section 3.1.1).17

Figure 1.1 shows the schematics of the MPECVD reactor. It comprises four main elements:

the microwave generator, the system for regulating the pressure, the gas supply system, and

Introduction

Chapter I Page 6

the vacuum chamber. A high frequency electromagnetic field pattern is created by the

microwaves and the reactant gases (CH4, H2 and trimethyl borane - TMB) are heated and

excited to form a plasma ball. The substrate sits just below the visible edge of the plasma ball,

on top of a molybdenum substrate holder. The reactors were also equipped with pyrometers to

monitor the temperature during growth.

Figure 1.1 Schematics of an MPECVD reactor

During growth, the boron atoms substitute the carbon atoms when the concentration is low,

and at higher concentration they occupy neutrally interstitial positions. In low doping cases,

the substituted boron atoms are bonded to neighboring carbon atoms in the sp3 configuration

and in the ground state the holes provided by boron atoms are bound to one of the three-fold

degenerate impurity states with a binding energy of 0.38 eV.18 At higher boron

concentrations, as the average distance between boron atoms is close to the acceptor Bohr

radius, metallic conduction appears at room temperature with conductivities of a few 100

ohm−1.cm−1. The growth conditions are summarized in table 1.2. All growths were performed

Introduction

Chapter I Page 7

using a low methane to hydrogen ratio ([CH4]/[H2] = 0.22%). TMB mixed with hydrogen at

2000 ppm is used as the boron gas source. These parameters were chosen based on the

previous studies conducted in the group.19 A recent study was conducted to clarify whether

diborane or TMB would be more appropriate for diamond growth. It appears that even in

single crystal diamond, TMB appears as efficient as diborane for boron doping.20

Table 1.2 Boron doped nanocrystalline layers synthesis conditions

Power 1.5 kW

Pressure 40 mbars

H2 88 sccm

CH4 0.22 sccm

TMB 12 (2000 ppm) sccm

Duration 10 - 15 hours

Temperature 750 °C

1.4 BDD film characterization

The synthesized BDD films were characterized using ellipsometry, Raman spectroscopy, X-

ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), seconday ion

mass spectroscopy (SIMS) cyclic voltammetry (CV) and electrochemical impedance

spectroscopy (EIS). These characterization techniques helped us to evaluate the continuity of

the film and the boron incorporation.19 Studies on synthesis of highly doped diamond

electrodes were already conducted in our lab based on different boron concentration.

Subsection 1.4.1 and 1.4.2 recalls the structural, morphological and electrochemical

characterization of the synthesized electrodes.19,21

1.4.1 Structural and morphological characterization

The thickness was characterized using a UV–Vis Spectroscopic Ellipsometer (Horiba Jobin-

Yvon, UVISEL) and the films were almost uniformly thick throughout the wafer and were

around 500 nm thick for a growth time of 15 hours. Secondary ion mass spectrometry (SIMS)

depth profiling was performed at CNRS GEMaC (Meudon) using a Cameca IMS 4f ion

microprobe. The measured boron concentration was approximately (2 × 1021 at.cm−3). The

Raman spectra were obtained using a Horiba Jobin–Yvon® SAS confocal Raman HR800.

Introduction

Chapter I Page 8

They were recorded on a cooled CCD detector after excitation of the films with the 325 nm

line of a He–Cd laser (spot size < 1 µm2).

Figure 1.2 Raman spectra of different BDD films at different boron concentrations.19

Figure 1.3 Scanning electron microscopic image of BDD film.

The UV excitation was chosen because the Raman shift of the diamond 1332 cm−1 signal

appears more clearly with this excitation due to a lower Fano deformation. Figure 1.2 shows a

Introduction

Chapter I Page 9

diamond peak position slightly below, but very close to 1332 cm−1, typical of boron

concentrations up to the metallic limit (3 × 1020 at.cm−3). The morphology and grain size were

assessed using SEM. The typical morphology of a boron-doped nanocrystalline layer (BNCD)

is shown in figure 1.3. The diamond crystals are highly faceted, with an average grain size of

100 nm and the obtained diamond films are uniform and continuous.

The surface analysis was performed using X-ray photoelectron spectroscopy. XPS analysis

was carried out on samples directly after the growth. The spectrometer consists of a

hemispherical analyzer and an Al Kg anode supplied with a monochromator. Binding

energies were referenced to the Au 4f 7/2 peak located at 84 eV.22 According to the

experimental geometry, the probed depth was estimated to be 1 nm. A curve-fitting procedure

was carried out to extract the components in the C1s spectra using Voigt functions with a

Lorentzian half-width of 0.2 eV. The Gaussian width was considered as an adjustable

parameter. Then the area of each peak was calculated and the ratio of this area was recorded

with respect to the total area.

Figure 1.4 XPS Survey of BNCD thin film.

0 200 400 600 800 1000

0

1000

2000

3000

4000

5000

6000

Inte

nsity (

a. u

.)

Binding energy (eV)

C 1s

Introduction

Chapter I Page 10

Figure 1.4 shows the XPS wide spectra of an ‘as grown’ sample. No oxygen was detected

using XPS corresponding to values below the detection threshold of 0.5 at %. In addition to

the C-C sp3 / C-H major peak located at 283.9 eV, the C1s signal exhibits a shoulder at + 0.6

eV corresponding to CHx (x>1) bonds which represents 19% of the C1s total area (Table 1.3).

Table 1.3 Percentages of total C1s area of the XPS components for ‘as grown’ BDD film

C-C sp2 C-C sp3 / C-H CHx C-OH C-O-C

‘As grown’ BDD < 0.5 % 81 % 19 % - -

1.4.2 Electrochemical characterization

An ideal electrode for bio-analysis need to possess wide working potential window to detect

species over a wide potential range, high signal-to-background ratio (low background current)

to have enhanced signal and sensitivity, low adsorption properties to reduce electrode fouling,

high electrochemical reactivity and high microstructural and morphological stability to

guaranty corrosion resistance. On the other hand, for neural recording and prosthesis high

impedance, low noise and corrosion resistive electrode is desirable. Refer to section 1.5 for

more details.

Two techniques: CV and EIS, were used to characterize the electrochemical (EC) properties

of the electrode.The BDD electrode was fixed on a non-reactive one-compartment cell with a

working surface area of typically 0.33 cm² and gallium-indium eutectic alloy was used

between the silicon substrate and the copper plate for the electrical contact as shown in figure

1.5.

When an electrode is in contact with an electrolyte, a double layer is formed at the

electrode/electrolyte interface as electric charges are accumulated on the electrode surfaces

and ions of opposite charge are arranged in the electrolyte side (figure 1.6). The electric

double layers formed at the electrode/electrolyte interfaces are accessible to ions present in

the electrolyte.23,24 The interactions between the ions in solution and the electrode surface are

assumed to be electrostatic in nature and result from the fact that the electrode holds a charge

density which arises from either an excess or deficiency of electrons at the electrode surface.

In order for the interface to remain neutral the charge held on the electrode is balanced by the

redistribution of ions close to the electrode surface. The Stern layer is defined as the compact

Introduction

Chapter I Page 11

layer of immobile ions strongly adsorbed to the electrode surface.25 Beyond the Stern layer is

the so-called diffuse layer where ions are mobile under the coupled influence of electrostatic

forces and diffusion. The result is analogous to an electrical capacitor which has two plates of

charge separated by some distance.26

Figure 1.5 Schematic diagram of 3 electrodes setup: WE (Working Electrode), RE

(Reference Electrode) and CE (Counter Electrode).

Figure 1.6 Schematics of electrical double layer at electrode/electrolyte interface.

The electrochemical window (potential range over which the electrolyte is not reduced or

oxidized at an electrode) in which electrode charging is capacitive is the double layer region

Introduction

Chapter I Page 12

and is determined by cyclic voltammetry.27 CV measurements consist of imposing an electric

potential at the electrodes which varies periodically and linearly with time. The resulting

electric current is recorded. The total charge accumulated at the electrode surface can be

found by integrating the electric current with respect to time.28

The EC characterizations were carried out using a 3 electrode setup. Ultrapure deionised (DI)

water (Millipore Direct Q3) was used to prepare all the solutions. During cyclic voltammetry

an Ag/AgCl/3M KCl reference electrode and a platinum mesh counter electrode were used

while BDD was the working electrode. Figure 1.7 shows a typical voltammogram of the BDD

electrode with a potential window greater than 3 V (in accordance with the literature values29)

in 0.5 M LiClO4 aqueous solution. As the potential window corresponds to the potential range

between which water gets electrolyzed, many electro-active species with high oxido-reduction

potential can be detected. The double layer capacitance (CD) probed for this electrode is 6.25

µF.cm-2. Lower the value, lower the signal-to-background ratio. The double layer capacitance

is up to one order of magnitude lower than that of glassy carbon electrode.29

Figure 1.7 Cyclic voltammogram of ‘as grown’ electrode in 0.5 M aqueous LiClO4 solution

scanned at 0.1 V s-1.

Introduction

Chapter I Page 13

Figure 1.8 Cyclic voltammogram of an ‘as grown’ BDD electrode in ferri/ferrocyanide

scanned at 0.1 V s-1.

Equimolar (1 mM) solutions of potassium ferricyanide(III)/ potassium

hexacyanoferrate(II).trihydrate (Acros Organics) were prepared in 0.5 M potassium chloride

(Acros Organics) solution. The CV of BDD in ferri/ferrocyanide redox couple yields faradaic

peaks and the peak separation Ep (= Epc - Epa) should be 58/n mV at all scan rates at 25 °C.

Figure 1.8 shows one such CV and the calculate Ep is 63 mV. Well defined peaks and peak

separation close to theoretical value indicate higher electrochemical reactivity.30 A high Ep

shows higher charge transfer resistance (resistance to the charge transferring chemical

reaction at the electrode-electrolyte interface).31 High electrode reactivity is essential for

electro-analytical application as it increases the sensitivity, signal-to-background ratio and

reproducibility of the measurement.

Impedance spectroscopy is a powerful tool to analyze the impedance response of an

electrochemical system and is sensitive to surface phenomena.32 An electrochemical system

can be represented by an equivalent circuit of lumped resistors and capacitors. A frequently

used circuit, called the Randles circuit, is shown in figure 1.9 a, is used to analyze the

impedance spectra, where RS is the solution resistance, RT the charge transfer resistance, CD

the double layer capacitance and W the Warburg diffusion impedance.

Introduction

Chapter I Page 14

During an impedance measurement, a frequency response analyzer (FRA) is used to impose

an AC signal of low amplitude on top of a DC potential. The frequency is varied from as high

as 105 Hz to as low as about 10-3 Hz in a set number of steps per decade of frequency.

Dividing the input voltage by the output current furnishes the impedance. The variation in

impedance (magnitude and phase angle) is used for the interpretation. The AC voltage and

current response of the cell is analyzed by the FRA to determine the resistive, capacitive and

inductive behavior of the cell at different frequencies.

In the equivalent circuit, the faradaic component arises from the electron transfer via an EC

reaction across the interface by overcoming an appropriate activation barrier (RT) along with

resistance represented by conductive pathways for ion (RS). The nonfaradaic current results

from charging the double-layer capacitor associated with space-charge polarization regions

(CD). The mass transport of the reactants and the products provides another class of

impedance: Warburg component (W).

EIS data for electrochemical cells such as fuel cells are most often represented by Nyquist

plot (figure 1.9 b). A complex plane or Nyquist plot depicts the imaginary impedance, which

is indicative of the capacitive character of the cell, versus the real impedance of the cell. The

semicircle relates to charge-transfer controlled process and its intercept with the X axis gives

RS and RT values (as seen in figure 1.9 b).

(a) (b)

Figure 1.9 (a) Randles equivalent circuit and (b) Nyquist Plot which describes impedance

behavior of a simple electrochemical cell

The electron transfer rate constant k0 (k0 is measured in 1mM ferro/ferricyanide in 0.5 M KCl,

throughout the report, unless stated) was experimentally determined by EIS over a frequency

Introduction

Chapter I Page 15

range of 50 kHz–1Hz with logarithmic point spacing and potential amplitude of 0.01 V rms

while the BDD was maintained at open circuit potential. The EIS was done in a three

electrode setup where BDD electrode is the working electrode, platinum wire the pseudo-

reference electrode and platinum mesh the counter electrode. The electrodes were rinsed in DI

water and dried under flow of argon gas prior to each experiment. The k0 value was

determined from the Nyquist plot fitted using ZSimWin 3.21 software. Unless stated

otherwise the potential is always given versus an Ag/AgCl reference electrode.

The quality of an electrode can be related to its electron transfer rate k0 which is defined by

equation (1.1) and used to characterize our electrodes in the rest of the document:

0T0 CnSFR

1

nF

RTk

(1.1)

Where R = Universal gas constant, T = Absolute temperature (K), S = Surface area of the

electrode (cm²), F = Faraday’s constant (96500 C.mol-1), RT = Electron transfer resistance of

electrode (ohm), C0 = Concentration of redox couple (mol.cm-3), n = number of electrons

transferred. The higher the value of k0, the better the reactivity of the electrode. The

experimental and fitted Nyquist plot of the electrode is depicted in figure 1.1 and k0 calculated

for this electrode was 0.5 cm.s-1.

Figure 1.10 Experimental and fitted Nyquist plot of an ‘as grown’ BDD electrode.

Introduction

Chapter I Page 16

1.5 BDD films for electrochemical sensors and biosensors

The synthesized BDD films exhibit excellent electrochemical characteristics. The objective of

the thesis is to realize electrochemical sensors for biomedical applications. In biomedical and

environmental applications, electrochemical sensors are generally well adapted due to their

good sensitivity, fast measuring time, portability, low power consumption and cost

effectiveness. BDD electrodes have been a major focus of research and development in

electrochemical and biomedical sensors due to their electro-analytically advantageous

features,5,29,33,34 namely: wide potential window in aqueous electrolytes (> 3 V), corrosion

stability in aggressive media, morphological and structural stability at very high current, low

background current and bio-inertness.

Owing to diamond’s stability and high activity to oxidize organic compounds,

electrochemical treatment of waste water provides an attractive alternative to traditional

methods. Compared to other electrode materials such as glassy carbon, metal oxide

electrodes, diamond has proven to be an excellent candidate because of its high anodic

stability and wide electrochemical window.35 Unlike conventional electrodes, diamond is less

susceptible to fouling and does not release toxic ions. The organic compounds are oxidized to

CO2 by electrogenerated hydroxyl radical (OH*). Organic pollutants oxidized on diamond

electrodes include, carboxylic acids, nitrophenols, phenols, chlorophenols, cyanides, etc.29

BDD electrodes are suitable mercury-free electrochemical sensors to detect trace metals such

as lead using anodic stripping voltammetry.36 Several techniques of anodic stripping

voltammetry have been successfully employed for lead detection which involves toxic

mercury film or mercury drop electrodes. BDD electrodes are ideal due to their enlarged

potential window, low background current and long term stability. Simultaneous

quantification of other heavy metals such as zinc, cadmium and copper is also possible using

BDD electrodes.37

Amperometric biosensors based on BDD electrodes have also attracted the interest of many

researchers as they combine the merits of biosensors such as specificity, sensitivity, and fast

response with the superior properties of BDD electrodes. Carbon-based materials such as

glassy carbon, porous carbon, carbon nanotubes and carbon nanofibers, etc. possess

advantages such as simple preparation technique, large potential ranges and ease of surface

modification. However, fouling of the electrode has to be dealt with and this leads to frequent

polishing or disposal of electrodes after limited usage. BDD electrodes are less subjected to

Introduction

Chapter I Page 17

fouling than other carbon-based electrodes.38 Other advantages include the low background

current increases the sensitivity when compared to the other electrodes.39 Development of

new chemical approaches to attach biomolecules covalently to diamond surfaces has led to an

entirely new class of electrochemical biosensors. The control of surface termination is crucial

for the detection of charged biomolecules. H-terminated BDD electrodes were used to detect

biomolecules such as glucose, DNA, NADH, etc.38,40,41 O-terminated BDD electrodes were

employed for detection of dopamine,42 tyrosinase modified boron doped diamond electrode to

detect Cresol,43 etc. A detailed example of biosensor developed using BDD electrodes to

detect uric acid in urine is explained in chapter 4.

Another interesting feature worth noting is the biocompatibility of diamond films.

Biocompatibility, by definition, is the quality of not producing toxic or injurious effects on

biological systems. Factors such as the chemical composition, roughness, surface energy,

topography and hydrophobicity of a material determine the interaction of cells and

biomolecules with the electrode.44 Specht et al. has demonstrated the biocompatibility of

diamond surfaces by selective attachment of mammalian neurons and ordered outgrowth of

neurites on its surfaces.45 As diamond is extremely inert, unlike metal implants, it does not

lead to cytotoxicity, allergy or malign effects caused by its degradation. Evaluation of

osteoblast and endothelial cell adhesion on diamond surfaces highlights not only the

mechanical and physical properties but also its attraction to cell colonization.46 Grausova et

al. has evaluated this possibility and recommends nanocrystalline diamond (NCD) as an

excellent candidate for in-vivo experiments and tissue engineering.46 Surface roughness of

biomaterials plays an important role in biocompatibility, especially on the adhesion of

endothelial cells. On the other hand, osteoblast cells are sensitive to the chemical composition

of the surface. Mitura et al. investigated on the impact of the contact of the NCD coating with

blood.47 NCD coating does not change in any significant way the activation of the elements of

coagulation of blood.

Biocompatibility, chemical stability, biostability and excellent mechanical characterization

indicate diamond’s potential for in-vivo and in-vitro biosensing applications. Some examples

of electrophysiological applications of diamond such as in-vitro neural recording and retinal

implants are described in detail in chapter 5.

Despite these promising properties when compared to classical electrodes, diamond electrodes

are also susceptible to fouling when used in organic or biological fluids.35,48,49 An in-situ

Introduction

Chapter I Page 18

regeneration approach was developed as part of this thesis in order to solve the problems

associated with fouling and is described in detail in Chapter 2.

Chapter 3 deals with the fabrication processes for microelectrodes, these being better suited

for biomedical applications when compared to macroelectrodes. Microelectrodes exhibit

significant advantages over macro-electrode systems, such as a decreased ohmic drop, a

hemispherical diffusion layer, which extends into the solution, rapid establishment of a

steady-state signal, and higher S/N ratio. Moreover, they require very small sample volumes 50.

Several multi-electrode arrays systems have been reported elsewhere, where all

microelectrodes are connected together to form one single probe 51,52. Those electrodes, when

designed appropriately, still behave as individual microelectrodes at suitable scan rates, but

since they are all interconnected they offer the additional advantages of an enhanced electrical

signal when compared to single electrodes. Boron doped diamond (BDD) materials exhibit

superior electrochemical properties over other conventional electrode materials including low

capacitive background currents, wide potential window in aqueous media and corrosion

resistance in harsh environments 29,33,53. Thus the advantage of using microelectrodes over

macro-electrodes has been further improved by combining their unique properties resulting

from geometrical characteristics with the excellent electrochemical properties of BDD

materials 54–57.

1.6 Theory of microelectrodes

Ultramicroelectrodes (UMEs) are electrodes with a critical size dimension below 25 たm.

UMEs exhibit different behavior from those of macroelectrodes when the critical diameter is

equal to or less than the diffusion layer.54 The diffusion of electro-active species occurs in two

dimensions: radially with respect to the axis of symmetry and normal to the plane of the

electrode. The current density is not uniform across the face of the disk, but is greater at the

edge, which offers the nearest point of arrival to electro-reactant drawn from a large

surrounding volume.

The faradaic current that flows at any time is a direct measure of the rate of the

electrochemical reaction taking place at the electrode32. This current depends upon two

phenomena:

Introduction

Chapter I Page 19

1) Charge transfer kinetics

2) Mass transport rate

Charge transfer kinetics is defined as the rate at which electrons are transferred across the

interface and the rate at which species move from the bulk of the solution to the electrode is

known as mass transport rate. There are three modes of mass transfer:

a. Diffusion: defined as the movement of any species under the influence of chemical

potential or a concentration gradient.

b. Migration: the movement of charged particles in an electric field.

c. Convection: movement of material contained within a volume element of stirred

(hydrodynamic) solution. It can be both natural and forced convection.

The experiments were designed such that the mass transport rate is only limited to the

contribution from diffusion. Fick’s first law quantifies the movement of a species (under

diffusion control) with respect to the distance x from an electrode with the flux J0. The

relation is given by:

待 噺 待 岫柱大轍柱淡 岻 (1.2)

Where D0 is the diffusion coefficient, C0 the concentration, and x the distance. In the above

equation (equation 1.2), the units of flux J0 are in moles.cm−2.s−1. The faradaic current can be

obtained by simplifying the above equation and is given as:

噺 待 岾柱大轍柱淡 峇 】掴退待 (1.3)

where n is the number of electrons transferred per molecule of reactant, F is Faraday’s

constant, and A is the surface area of the electrode. Fick’s second law describes the time-

dependent changes in the concentration of the substance amount caused by the flux.

柱大轍柱担 噺 待岫柱鉄大轍柱淡鉄 岻 (1.4)

where ‘t’ is duration of experiment.

The diffusion layer is the region in the vicinity of an electrode where the concentrations are

different from their value in the bulk solution. The flux of the substance toward the electrode

is then described by the product of the diffusion coefficient of the substance, D0, and its bulk

concentration divided by the diffusion layer thickness, d. This quantity is defined by the

relationship:

Introduction

Chapter I Page 20

噺 岫に 待 岻待┻泰 (1.5)

When d << r (radius r of the electrode), the perturbation of the linear diffusion flux caused by

the hemispherical diffusion at the electrode edges extends to a short distance of several d from

the edge. The behavior of the electrode can be approximated by the infinite electrode model

as only a small part of the electrode surface is affected. When d >> r, the edge effect plays an

important role as the diffusional flux towards the electrode is constant with time but

inhomogeneous over the electrode surface. The diffusion layer is hemispherical in shape and

extends out into the solution. Figure 1.11 demonstrates the planar and radial diffusion profile

at a macro and micro electrode respectively. Under such conditions, the concentration of the

electro-active substance attains the limiting value described by the equation for steady-state

transport, which corresponds to the time derivative being zero in Fick’s second law and the

current goes to a steady state value.

(a)

(b)

Figure 1.11 Schematics illustration of (a) linear diffusion at a planar electrode and (b) radial

diffusion at a microelectrode.

The steady state limiting current is different for different types of ultramicroelectrodes based

on their geometry:32

(a) Spherical or hemispherical UME: 狸辿鱈 噺 ねぱ 待 待 (1.6)

(b) Cylindrical UME: 狸辿鱈 噺 に 待 待【 ぷ (1.7)

where 酵 噺 ね 待 【 態

Introduction

Chapter I Page 21

(c) Band UME: 狸辿鱈 噺 にぱ 待 待【 岫はね 待 【 態岻 (1.8)

where w = width of the band

(d) Disk UME: 狸辿鱈 噺 ね 待 待 (1.9)

Conclusion

In the first chapter, different diamond growth techniques and characterization methods are

described. The diamond electrodes were synthesized using MPECVD technique from a

mixture of methane, hydrogen and TMB. The physical and chemical properties of the thin

film were assessed using Raman spectroscopy, XPS, SEM, CV and EIS. Highly doped

diamond electrodes give rise to extreme reactivity and they are ideal candidates for electro-

analytical and biosensing applications. Chapter 2 addresses the problem associated with loss

of reactivity especially when the electrode is used in a biological medium. A novel activation

process is described in detail where diamond electrodes can be reactivated within the

biological fluid.

Introduction

Chapter I Page 22

Bibliography

(1) Jee, A.; Lee, M. Current Applied Physics 2009, 9, e144–e147.

(2) Tolbert, L. M.; Ozpineci, B.; Islam, S. K.; Chinthavali, M. S. From Proceeding (379) Power and Energy Systems 2003, 1–5.

(3) Neves, A. J.; Nazaré, M. H. Properties, Growth and Applications of Diamond; IET, 2001; p. 427.

(4) Ferro, S. Journal of Materials Chemistry 2002, 12, 2843–2855.

(5) Xu, J.; Granger, M. C.; Chen, Q.; Strojek, J. W.; Lister, T. E.; Swain, G. M. Analytical Chemistry News & Features 1997, 591 A – 597 A.

(6) Lee, S.; Lin, Z.; Jiang, X. Material Science and Engineering 1999, 25, 123–154.

(7) Matsumoto, S.; Sato, Y.; Tsutsumi, M.; Setaka, N. Journal of Materials Science 1982, 17, 3106–3112.

(8) Tang, Y.; Aslam, D. M. Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures 2005, 23, 1088.

(9) Mitsuda, Y.; Kojima, Y.; Yoshida, T.; Akashi, K. Journal of Materials Science 1987, 22, 1557–1562.

(10) Yugo, S.; Kanai, T.; Kimura, T.; Muto, T. Applied Physics Letters 1991, 58, 1036.

(11) Tsuda, M.; Nakajima, M.; Oikawa, S. Journal of American Chemical Society 1986, 108, 5780–5783.

(12) May, P. W.; Smith, J. a.; Mankelevich, Y. a. Diamond and Related Materials 2006, 15, 345–352.

(13) Butler, J. E.; Sumant, A. V. Chemical Vapor Deposition 2008, 14, 145–160.

(14) Fujimori, N.; Imai, T.; Doi, A. Vacuum 1986, 36, 99–102.

(15) Okano, K.; Kiyota, H.; Iwasaki, T.; Nakamura, Y.; Akiba, Y.; Kurosu, T.; Iida, M.; Nakamura, T. Applied Physics A Solids and Surfaces 1990, 51, 344–346.

(16) Koizumi, S.; Kamo, M.; Sato, Y.; Mita, S.; Sawabe, A.; Reznik, A.; Uzan-Saguy, C.; Kalish, R. Diamond and Related Materials 1998, 7, 540–544.

(17) Scorsone, E.; Saada, S.; Arnault, J. C.; Bergonzo, P. Journal of Applied Physics 2009, 106, 014908.

Introduction

Chapter I Page 23

(18) Wu, D.; Wang, Z. L.; Luo, Q.; Gu, C. Z.; Wang, N. L.; Li, C. Y.; Lu, X. Y.; Jin, Z. S. Physical Review B 2006, 73, 1–5.

(19) Vanhove, E.; de Sanoit, J.; Mailley, P.; Pinault, M. -a.; Jomard, F.; Bergonzo, P. Physica Status Solidi (a) 2009, 206, 2063–2069.

(20) Volpe, P.-N.; Arnault, J.-C.; Tranchant, N.; Chicot, G.; Pernot, J.; Jomard, F.; Bergonzo, P. Diamond and Related Materials 2012, 22, 136–141.

(21) Vanhove, E.; de Sanoit, J.; Arnault, J. C.; Saada, S.; Mer, C.; Mailley, P.; Bergonzo, P.; Nesladek, M. Physica Status Solidi (a) 2007, 204, 2931–2939.

(22) Seah, M. P. Surface and Interface Analysis 1993, 20, 243–266.

(23) Yoon, S.; Lee, J.; Hyeon, T.; Oh, S. M. Journal of The 2000, 147, 2507.

(24) Wang, H.; Pilon, L. Electrochimica Acta 2012, 64, 130–139.

(25) Birdi, K. S. Introduction to Electrical Interfacial Phenomena; Taylor & Francis, 2009; p. 167.

(26) Lvovich, V. F. Impedance Spectroscopy: Applications to Electrochemical and Dielectric Phenomena (Google eBook); John Wiley & Sons, 2012; p. 368.

(27) Green Industrial Applications of Ionic Liquids (Google eBook); Springer, 2003; Vol. 0, p. 584.

(28) Compton, R. G.; Banks, C. E. Understanding Voltammetry; World Scientific, 2011; p. 429.

(29) Panizza, M.; Cerisola, G. Electrochimica Acta 2005, 51, 191–199.

(30) Wang, W. L.; Jin, E. M.; Gu, H.-B. Transactions on Electrical and Electronic Materials 2012, 13, 121–124.

(31) Nicholson, R. S. Analytical chemistry 1965, 37, 1351–1355.

(32) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications; Wiley, 2001; Vol. 6.

(33) Swain, G. M. Analytical chemistry 1993, 65, 345–351.

(34) Granger, M. C.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.; Koppang, M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M. Anal. Chem. 2000, 72, 3793–3804.

(35) Iniesta, J.; Michaud, P. A.; Panizza, M.; Cerisola, G.; Aldaz, A.; Comninellis, C. Electrochimica Acta 2001, 46, 3573–3578.

Introduction

Chapter I Page 24

(36) Dragoe, D.; Sp<taru, N.; Kawasaki, R.; Manivannan, A.; Sp<taru, T.; Tryk, D. a.; Fujishima, A. Electrochimica Acta 2006, 51, 2437–2441.

(37) El Tall, O.; Jaffrezic-Renault, N.; Sigaud, M.; Vittori, O. Electroanalysis 2007, 19, 1152–1159.

(38) Rao, T. N.; Yagi, I.; Miwa, T.; Tryk, D. a; Fujishima, a Analytical chemistry 1999, 71, 2506–11.

(39) Zhou, Y.; Zhi, J. Talanta 2009, 79, 1189–96.

(40) Lee, J.; Park, S.-M. Analytica Chimica Acta 2005, 545, 27–32.

(41) Weng, J.; Zhang, J.; Li, H.; Sun, L.; Lin, C.; Zhang, Q. Analytical chemistry 2008, 80, 7075–83.

(42) Tryk, D. A.; Tachibana, H.; Inoue, H.; Fujishima, A. Diamond and Related Materials 2007, 16, 881–887.

(43) Zhou, Y.; Zhi, J. Electrochemistry Communications 2006, 8, 1811–1816.

(44) Ponsonnet, L.; Reybier, K.; Jaffrezic, N.; Comte, V.; Lagneau, C.; Lissac, M.; Martelet, C. Materials Science and Engineering: C 2003, 23, 551–560.

(45) Specht, C. G.; Williams, O. a; Jackman, R. B.; Schoepfer, R. Biomaterials 2004, 25, 4073–8.

(46) Grausova, L.; Kromka, a.; Bacakova, L.; Potocky, S.; Vanecek, M.; Lisa, V. Diamond and Related Materials 2008, 17, 1405–1409.

(47) Mitura, S.; Mitura, A.; Niedzielski, P.; Couvrat, P. Chaos, Solitons & Fractals 1999, 10, 2165–2176.

(48) Montilla, F.; Michaud, P. A.; Morallon, E.; Vazquez, J. L.; Comninellis, C. Electrochimica Acta 2002, 47, 3509–3513.

(49) Panizza, M.; Michaud, P. a.; Cerisola, G.; Comninellis, C. Journal of Electroanalytical Chemistry 2001, 507, 206–214.

(50) Stulik, K.; Amatore, C.; Holub, K.; Marecek, V.; Kutner, W. Pure and Applied Chemistry 2000, 72, 1483–1492.

(51) Niwa, O.; Xu, Y.; Halsall, H. B.; Heineman, W. R. Analytical chemistry 1993, 65, 1559–63.

(52) Kounaves, S. P.; Deng, W.; Hallock, P. R. October 1994, 66, 418–423.

(53) Yano, T.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. Journal of The Electrochemical Society 1998, 145, 1870–1876.

Introduction

Chapter I Page 25

(54) Soh, K. L.; Kang, W. P.; Davidson, J. L.; Basu, S.; Wong, Y. M.; Cliffel, D. E.; Bonds, A. B.; Swain, G. M. Diamond and Related Materials 2004, 13, 2009–2015.

(55) Lawrence, N. S.; Pagels, M.; Meredith, A.; Jones, T. G. J.; Hall, C. E.; Pickles, C. S. J.; Godfried, H. P.; Banks, C. E.; Compton, R. G.; Jiang, L. Talanta 2006, 69, 829–834.

(56) Tian, R.; Zhi, J. Electrochemistry Communications 2007, 9, 1120–1126.

(57) Pagels, M.; Hall, C. E.; Lawrence, N. S.; Meredith, A.; Jones, T. G. J.; Godfried, H. P.; Pickles, C. S. J.; Wilman, J.; Banks, C. E.; Compton, R. G.; Jiang, L. Analytical chemistry 2005, 77, 3705–8.

Introduction

Chapter I Page 26

CHAPTER II

Electrochemical activation of diamond electrodes

Electrochemical activation of diamond electrodes

Chapter II Page 28

Electrochemical activation of diamond electrodes

Chapter II Page 29

Introduction

Boron-doped diamond electrodes are extremely promising for a range of biomedical

applications as they exhibit a unique combination of favorable properties. Despite these

promising properties when compared to classical electrodes, unfortunately they too are

susceptible to fouling when used in organic or biological fluids.1–3 Electrode fouling can be

due to adsorption or adhesion of biomolecules such as proteins, enzymes, cells, intermediate

products of oxidation of organic compounds, etc.4–7 ‘As grown’ BDD electrodes are

hydrogen-terminated and they exhibit very good electrochemical reactivity characterized by

their high electron transfer rate k0. But ageing in air or in aqueous solution reduces the k0

value to considerably,8,9 possibly due to surface modification. The H-terminated, ‘as grown’

BDD electrode is gradually modified to O-terminated on exposure to air. Both fouling and

ageing affects the accuracy, sensitivity and reproducibility of the measurement and lifetime of

the electrode.

One of the thesis’ objectives was to develop a simple and fast electrochemical (EC) treatment

that can be used to retrieve the lost reactivity, thereby enhancing electrodes’ reusability over

long periods of measurement without degradation of the signal. This chapter describes a novel

EC activation process where a train of short current pulses is applied to the BDD electrodes.

The surface modification of EC-activated electrodes is investigated using X-ray Photoelectron

Spectroscopy and electrochemical characterization. Influence of current density, pH, and

number of pulses on EC-activation is discussed. Some examples of activation within

biological and synthetic fluids are also presented.

2.1 Ageing and fouling of the electrode

Two phenomena: electrode ageing and fouling, that affect the electrode reactivity and thereby

affect the accuracy of measurement are explained in detail with the help of cyclic

voltammetry (CV) and electrochemical impedance spectroscopy (EIS). The details of

electrochemical characterization were explained in chapter 1 (section 1.4.2).

2.1.1 Electrode ageing

BDD Electrodes, when fabricated from recently grown material, exhibit very high reactivity

with k0 values above 0.01 cm.s-1. The peak to peak separation of the oxidation and reduction

potential (∆Ep in [Fe(CN)6]3-/4- solution) of such an electrode was probed at 60 mV at a 100

Electrochemical activation of diamond electrodes

Chapter II Page 30

mV.s-1 scan rate (as shown in figure 2.1), a value close to the theoretical value for this couple,

thus demonstrating the extreme reactivity of the diamond surface. The k0 value for this

electrode was estimated to be 5.9 x 10-2 cm.s-1. However, when exposed to air for 30 days

(under laboratory conditions), the same electrode evidenced a decrease in the k0 value of

1100 % and ∆Ep decreased by 42% due to this ageing. Figure 2.2 shows a comparison

between the Nyquist plots of ‘as grown’ BDD electrode and the same electrode after ageing

where the transfer resistance RT (related to the diameter of the high frequency arc) increased

from 10 ohm to 105 ohm.

Figure 2.1 Cyclic voltammogram of ‘as grown’ electrode (solid line) and the same electrode

after 30 days exposure to air (dashed line) scanned at 100 mV s-1 in 0.5 M KCl solution

containing 1 mM [Fe(CN)6]3-/4-.

There have been several proposals suggested by various groups in order to establish a

relationship between the ageing and the electrochemical properties of BDD electrodes. Duo et

al. proposed that a non-negligible amount of sp2 species might be present in the grain

boundary which could be the reason for the high reactivity of ‘as grown’ electrode.10 A

relatively mild anodic polarization process was suポcient to transform the surface from

Electrochemical activation of diamond electrodes

Chapter II Page 31

hydrophobic to hydrophilic without changes in the crystal shape and size. The variation of

BDD electrode surfance is related to the oxidation of the sp2 states present on the surface

leading to a “non-active” polycrystalline diamond surface. Another study of this matter

indicated that the changes in the BDD electrodes during the anodic process are related to the

variation in the concentration of H at the surface.10 An Oxygen-terminated surface may be the

cause of the observed high surface electrical resistance, indicating the attainment of positive

electron affinity.11

A study has been conducted in our laboratory to address the evolution of H-terminated BDD

electrochemical properties with time.8 It was observed that these electrochemical evolutions

were due to modification of surface termination and partial inactivation. According to this

study, the XPS analysis of ‘as grown’ BDD electrode shows C-C sp3 and C-H (74% of total

area) and C-Hx (26%) peaks. After exposure to air, an increase in CHx bonds and a weak

contribution of C-OH bond was observed. Hence the electrochemical evolution of ‘as grown’

BDD electrode might be due to an increase in polyhydride carbon species and surface

oxygenation.

Figure 2.2 Comparison between the Nyquist plot of ‘as grown’ BDD electrode and that of the

same electrode which has been exposed to air for 30 days.

Electrochemical activation of diamond electrodes

Chapter II Page 32

2.1.2 Electrode fouling

The effects of fouling are much more severe than those associated with electrode ageing as k0

in the former case is found to decrease by several tens of times after using the electrode in a

biological medium. In order to demonstrate the effect of fouling, a set of 5 CV cycles from -

0.4 V to 1.1 V vs Ag/AgCl at 100 mV s-1 was performed in human urine using a freshly

prepared BDD electrode. After each trial EIS was carried out to assess the k0 value. An

oxidation peak (P1) was observed on the CV at approximately 0.5 V vs Ag/AgCl (figure 2.3).

The peak P1 corresponds to the oxidation potential of uric acid and ascorbic acid in urine.12

Chapter 4 is dedicated to the quantification process of uric acid in presence of ascorbic acid in

human urine.

The amplitude of P1 is observed to significantly decrease after each trial. The attenuation of

current is not because of diffusion-limited phenomena as the electrode was washed

thoroughly in DI water and dried before each CV. Also the oxidation potential of this peak

has shifted towards more positive potentials and the k0 value decreases after each trial (figure

2.3 and table 2.1). Before the measurement, a k0 value of 0.27 cm.s-1 was measured, whereas

after five trials it decreased to less than 10-3 cm.s-1. This clearly illustrates the effect of the

fouling of the electrode.13 Human urine contains many proteins such as macroglobulin,

fibrinogen, antegens of epithelial cells, etc.14 in addition to carbohydrates, hormones, fatty

acids and other organic and inorganic compounds. When human urine is not diluted and used

as the analyte these complex molecules get adsorbed on the electrodes surface thereby

blocking the electron transport and reducing the electron transfer rate to considerable extend.

Table 2.1 Electron transfer rate (k0) of ‘as grown’ electrode and values after subsequent scan

in human urine.

k0 (cm.s-1)

As grown 0.27

After scan 1 0.011

After scan 2 0.005

After scan 3 0.003

After scan 4 0.002

After scan 5 0.001

Electrochemical activation of diamond electrodes

Chapter II Page 33

Figure 2.3 Cyclic voltammogram in human urine from -0.4 V to 1.1 V at 100 mV s-1 where J

is the current density in µA.cm-2 and E is the voltage in volts. The electrode was cleaned

thoroughly in deionized water prior to each scan and hence the attenuation of the peak (P1) is

due to fouling and not because the solution surrounding electrode is depleted of electro-active

species.

2.2 Activation process

Several approaches have been investigated to overcome the issue of fouling and/or ageing.

For instance, coating the electrochemical sensor with chemically-inert polymers such as

Nafion® enhances the antifouling capability.6 Surfactant-modified electrodes also enhance

the resistance to protein adsorption and cell adhesion. However, polymer or membrane

deposition increases the degree of complexity of fabrication of the electrochemical sensor,

decreases the electrode reactivity and, as a result, the lifetime of the modified electrode will

be shorter than that of an ‘as grown’ electrode. Alternatively, the sono-electrochemical

method provides in-situ cleaning accompanied with electrochemical measurement.4

Nevertheless the power consumption, accuracy of measurement and simplicity of the design

Peak P1

Electrochemical activation of diamond electrodes

Chapter II Page 34

of the sensor have, in this case, to be compromised. Hydrogen plasma treatment is another

approach that leads to clean H-terminated BDD surface.15,16 However, the use of plasma

treatment is clearly not practical, especially when measurements have to be performed outside

the laboratory.

Electrochemical techniques have also been developed to activate an aged or fouled electrode.

In particular, an aged BDD electrode can be reactivated using cathodic pre-treatment by

applying , typically, -3 V for 3 to 30 minutes in 0.5 M H2SO4 aqueous solution.9 Also the non-

diamond sp² impurities can be eliminated by anodic treatment in aqueous electrolyte.15

Electrochemical reactivity of an aged electrode can be improved by performing 10 redox

cycles in 0.5 M LiClO4 aqueous electrolyte (de-aerated) from 400 µA.cm-2 to -400 µA.cm-2 at

100 mV s-1.8 Rodrigo et al. have demonstrated that anodic treatment of a fouled electrode at

10 mA.cm-2 for 30 minutes can enable one to recover the initial reactivity of the electrode.17

This is due to oxidation of the organic molecule to CO2 by the electrogenerated hydroxyl

radical (OH*). A pulsed cleaning technique has been reported by Mahé et al.18 where

alternating current pulses of amplitude ± 250 mA.cm-2 in 1 M HNO3 were applied to clean the

graphitic domains on diamond electrodes. The total activation time was, however, 400

seconds.

We have developed an improved EC activation process for BDD electrodes, based on the

application of specific current or potential pulses, that leads to electrochemical cleaning and

to the recovery of remarkable electron transfer rates (k0 value above 10-3 cm.s-1) with good

stability and an activation time shorter than known in the prior art (as short as 200 ms). Based

on the activation time, pulse duration and pulse amplitude, two protocols were empirically

derived: ‘standard activation protocol’ and ‘standard cathodic activation protocol’. The details

are explained in section 2.4. The standard activation protocol consists of a pulse train of 50

current pulses of alternating amplitude (10 mA.cm-2 and -10 mA.cm-2) and of equal duration

(100 ms) applied between working and counter electrodes in 0.5 M aqueous LiClO4 solution.

The ‘standard cathodic activation protocol’ consists of a train of 50 current pulses of

amplitude -20 mA.cm-2, duration of 110 ms and a duty cycle of 91%. The optimal values

required depend on the electrolyte used and were therefore modified accordingly for other

electrolytes.

Electrochemical activation of diamond electrodes

Chapter II Page 35

This innovative EC treatment was tested on the aged electrode by using the ‘standard

activation protocol’ to activate the electrode. It was observed that the electrode activity was

recovered, with measured ∆Ep values brought back to 60 mV and k0 values reaching 6.1x 10-2

cm.s-1 after this novel EC treatment. The time required to bring back the lost reactivity of the

aged electrode, in this case, was as low as 10 seconds. The ∆Ep and k0 values of ‘as grown’,

aged and activated electrode are summarized in table 2.2. With respect to other techniques as

previously reported in the literature, this technique is much faster, can be performed using an

extremely simple electronic setup (pulse generator), and requires lower power resources.

Table 2.2 Comparison of the peak-to-peak voltage separation (∆Ep) and electron transfer rate

(k0) values of ‘as grown’, aged and activated electrode.

〉EP (mV) k0 (cm.s-1)

‘as grown’ 60 5.94 x 10-2

Aged 85 5.04 x 10-3

After Activation 60 6.05 x 10-2

Similarly, the electrode fouled in human urine was treated with the EC activation technique

using the following parameters: a set of 150 current pulses of alternating amplitudes (10

mA.cm-2 and -10 mA.cm-2) and of equal duration (100 ms) was applied to the working

electrode with respect to the counter electrode in 0.5 M aqueous LiClO4 solution. The total

number of pulses was tripled when compared to the ‘standard activation protocol’ because it

was observed that the k0 value after 50 pulses was not as high as that of the ‘as grown’

electrode (although significantly higher than 0.01 cm.s-1). This corresponds to an overall

activation time of 30 seconds.

The results were extremely encouraging as it can be observed in figure 2.4 that the CV curve

after activation almost perfectly coincides with that of the first scan in urine. The difference in

amplitude of the P1 peak for these two curves remained below 0.01% and the k0 value was

brought back to more than 0.2 cm.s-1 by this activation. Activation process also shifted the

peak P1 back to 0.5 V vs Ag/AgCl which was initially shifted towards more positive

potentials due to fouling. During the activation, adsorbed species were oxidized to carbon

dioxide by the hydroxyl radicals that were electro-generated during pulsing and/or desorbed

Electrochemical activation of diamond electrodes

Chapter II Page 36

by the gaseous species. The explanation of the activation process is described in detail in

section 2.5.

Figure 2.4 Comparison of the cyclic voltammogram of an ‘as grown’ electrode (solid line)

and the same electrode after activation (dotted) where J is the current density in µA.cm-2 and

E is the voltage in volts. The electrolyte is human urine and the scan rate is 100 mVs-1.

2.3 Activation in other synthetic electrolytes

The aforementioned EC activation approach can also successfully be performed in most

inorganic salt solutions as well as in organic solutions. In order to demonstrate this, electrodes

were fouled and then activated in various electrolytes. For example, the BDD electrodes were

fouled in red wine (Chassagne-Montrachet 1er Cru) by maintaining them at 0.7 V vs Ag/AgCl

for 20 s. The red wine sample was particularly chosen since it led to high levels of electrode

fouling. The fouled electrodes exhibited k0 values below 10-3 cm.s-1. The electrodes were then

activated in different salt solutions of 0.2 M using 8 alternating current pulses (+1.5 mA.cm-2

and -1.5 mA.cm-2) of 2 seconds each. The k0 values after activation in each salt solution is

Electrochemical activation of diamond electrodes

Chapter II Page 37

summarized in table 2.3. This demonstrated the role of cathodic current pulses in electrode

cleaning.

For non-electroactive electrolytes such as LiClO4 and Na2SO4 at this current density, the

negative potential goes beyond 1.6 V where H2 gas is electro-generated. The electro-

generated H2 gas helps in desorption of adsorbed organic species and H-termination of the

electrode. Electrochemical hydrogenation of the surface is due to attachment via a production

of hydrogen radicals at negative voltages.19 Whereas for the electroactive electrolyte the

current generated was mainly due to metal deposition rather than H2 electro-generation and

hence the k0 value is lower. At a relatively low current density of 1.5 mA.cm-2, electroactive

species such as Cu, Zn and Mn ions are electrodeposited on the surface and thereby decrease

the yield of OH*, which also plays a critical role in electrode cleaning.

Table 2.3 The electron transfer rate (k0) value determined from the Nyquist plot after

activation in a range of salt solutions.

Salt Solvent k0 (cm.s-1)

LiClO4 Water 0.0268±8x10-4

Na2SO4 Water 0.0205±6x10-4

MnSO4 Water 0.0108±0.004

ZnSO4 Water 0.0027±7x10-4

CuSO4 Water 0.0013±6x10-4

H2SO4 Water 0.0197 ±3x10-3

TBATFB Acetonitrile 0.0095±4x10-4

The deposition of metal was observed by the anodic stripping voltammogram where Zn was

stripped at -0.9 V, Cu at 0.4 V versus Ag/AgCl and for Mn two peaks were observed

(oxidation and reduction peaks). Figure 2.5 shows the anodic stripping voltammogram

indicating that at low current densities metal ions were deposited on the surface. It was

observed that at higher current densities (> 10 mA.cm-2), the k0 value was higher after

activation in some electro-active electrolyte because the yield of H2 generation and H

termination increases with current density. Activation process in 0.2 M tetrabutylammonium

tetrafluoroborate (TBATFB) in acetonitrile solution demonstrated that the activation process

Electrochemical activation of diamond electrodes

Chapter II Page 38

can also be successfully performed in organic solvents to achieve considerably high k0 values

above 10-3 cm.s-1.

Figure 2.5 Stripping voltammogram for copper, manganese and zinc. The electrolyte solution

contained 0.2 moles of either Cu2+, Mn2+ or Zn2+ ions.

2.4 Influence of pH, current density and number of pulses on activation

To probe the influence of factors such as the amplitude of the current density, the pH of the

electrolytic solution and the number of pulses in a pulse train, one such parameter was varied

at a time while the two others were kept constant. To understand the influence of pH and

number of pulses, the total activation duration was fixed at 16 seconds. To analyze the

influence of current density on activation, the duration of activation was limited to 100 ms

because the current density was varied from 1 µA.cm-2 to 100 mA.cm-2. 16 seconds of

activation at 100 mA.cm-2 electrolyzes the solution rapidly and may also affect the surface

termination although BDD is known to be robust. At first, the electrodes were fouled in the

Electrochemical activation of diamond electrodes

Chapter II Page 39

red wine to reach k0 values below 10-3 cm.s-1 and then activated in 0.5 M aqueous LiClO4

solutions with different parameters.

2.4.1 pH vs activation

After fouling, the electrodes were activated in 0.5 M aqueous LiClO4 solutions with varying

pH values. The pH was adjusted to 1.5, 4.5, 7, 9.5 and 12.5 by adding either H2SO4 or NaOH.

Activation was performed using 8 current pulses of alternating amplitude (1.5 mA.cm-2 and -

1.5 mA.cm-2) and duration of 2 seconds each. The more the alkaline the solution, the better

the activation of the electrode as displayed in figure 2.6. The k0 value of the electrode after

activation in a solution of pH 1.5 was found to be 1.34 x 10-2 cm.s-1 whereas that of an

electrode activated in alkaline solution of pH 12.5 was 4.82 x 10-2 cm.s-1.

Figure 2.6 pH dependence of the activation process. The higher the pH, the better the

activation as well as the value of electron transfer rate (measured in 1mM ferro/ferricyanide in

0.5 M KCl) k0 measured in cm.s-1.

We observed an increase of around 250% in k0 when the fouled electrode is activated in an

alkaline medium as compared to that in an acidic medium. This could possibly be associated

with the fact that, when alternating positive and negative pulses were applied through the

Electrochemical activation of diamond electrodes

Chapter II Page 40

working electrode, OH* radicals generated may play a relevant role in oxidizing organic

compounds.20,21 Electrode fouling is associated with the deposition or adsorption of organic

compounds on the electrode surface and the OH* radicals oxidize them to CO2. Alkaline

solution produce more OH* radicals when compared to acidic solutions.

2.4.2 Current density vs activation

Increasing the current density of the current pulses causes a very dramatic increase in k0

values (figure 2.7). The electrodes were fouled in red wine and then were activated in 0.5 M

aqueous LiClO4 solution using 4 alternating current pulses of 100 ms duration each, while the

pH of the solution was kept constant (pH = 4.5). This activation was performed using varying

current density amplitudes of 1 µA.cm-2, 10 µA.cm-2, 100 µA.cm-2, 1 mA.cm-2, 10 mA.cm-2

and 100 mA.cm-2, respectively. The k0 of the electrode after activation using a current density

of 1 µA.cm-2 was 6.97 ± 0.8 x 10-4 cm.s-1 and that of an electrode activated using a current

density of 100 mA.cm-2 was 1.71 ± 0.3 x 10-2 cm.s-1. Maintaining both the total time of

activation and pH constant, the result is an increase in the k0 value by up to a factor 25 with

the current density.

Figure 2.7 Effect of current density on the activation process. The higher the current density,

the higher the electron transfer rate k0 (cm.s-1), where J is the current density in µA.cm-2.

Electrochemical activation of diamond electrodes

Chapter II Page 41

2.4.3 Number of pulses vs activation

The number of pulses required to activate the electrode also plays a very critical role. This

was assessed by varying the total number of pulses per activation while the total activation

time, pH and current density were kept constant (pH = 4.5). Alternating current pulses with

absolute amplitude of 1.5 mA.cm-2 were applied through the fouled electrode in 0.5 M LiClO4

solution and the total activation time was 16 seconds. The fouled electrode was activated

using series of 2, 4, 8, 16, 32 and 64 alternating pulses and with the corresponding pulses

exhibiting durations of 8, 4, 2, 1, 0.5 and 0.25 seconds, respectively. It was observed that the

higher the number of pulses for a finite activation time, the better the yield of activation.

When two alternating pulses of 8 seconds each were applied, the post activation k0 was 8.2 ±

0.7 x 10-3 cm.s-1, and values reached 2.1 ± 0.1 x 10-2 cm.s-1 for k0 when 64 alternating pulses

were applied (figure 2.8). A 250% increase was observed for the k0 value when the number of

pulses was increased from 2 to 64 while the total activation time was kept constant for both.

This justifies using several pulses instead of just 2 pulses (alternating cathodic and anodic

pulses).

Figure 2.8 Impact of the number of pulses on the activation process, where k0 is electron

transfer rate (cm.s-1).

Electrochemical activation of diamond electrodes

Chapter II Page 42

From figure 2.8 it is clear that there is a large increase in k0 value when the number of pulses

was increased from 1 to 10 and the slope tends to level off when the number of pulses was

increased further. Similarly, there is a large increase in k0 value when the current density was

increased beyond 10 mA.cm-2 as seen in figure 2.7. Based on these observations it is

suggested that the ‘standard activation protocol’ successfully cleans the electrode in a non-

reactive aqueous electrolyte: Positive and negative trains of pulses of the same amplitude (±

10 mA.cm-2) and duration (100 ms each) and series of 100 such pulses.

2.5 Surface analysis and activation mechanism

The surface analysis was performed using X-ray Photoelectron Spectroscopy (see chapter 1,

section 1.4.1 for more details). XPS analysis was carried out on two ‘as grown’ electrodes:

one sample was analyzed directly after growth and the other sample was activated using

‘standard activation protocol’ prior to the XPS analysis. For the ‘as grown electrode’ no

oxygen was detected using XPS corresponding to values below the detection threshold of 0.5

at %. In addition to the C-C sp3 / C-H major peak located at 283.9 eV, the C1s signal exhibits

a shoulder at + 0.6 eV corresponding to CHx (x>1) bonds which represent 19% of the C1s

total area (table 2.4). This signature was previously reported for hydrogenated diamond

surfaces.8 Finally, a negligible contribution (< 0.5 at %) was detected at 282.5 eV

corresponding to sp2 carbon. This is related to the grain size of nanocrystalline boron-doped

diamond films leading to a weak contribution from grain boundaries.

Table 2.4 Percentages of total C1s area of the XPS components for ‘as grown’ electrode and

electrode after EC activation.

C-C sp2 C-C sp3 / C-H CHx C-OH C-O-C

‘As grown’ BDD < 0.5 % 81 % 19 % - -

Activated BDD < 0.5 % 72 % 20 % 6 % 2 %

After activation, the oxygen concentration extracted from the O1s core level was 3.8 at %.

Two new contributions have to be taken into account at the C1s core level. The first one is

located at + 1.3 eV from the C-C sp3 / C-H peak and could be assigned to C-OH bonds. The

second weaker contribution at 1.9 eV from the C-C sp3 is attributed to C-O-C bonds. The

respective area ratios are given in table 2.4. This surface chemistry is close to the one

Electrochemical activation of diamond electrodes

Chapter II Page 43

measured after activation in LiClO4 electrolyte where C-OH (4 %) and C-O-C (5 %)

contributions were also present.8 Figure 1.4 and 2.9 shows the XPS spectra of ‘as grown’ and

activated BDD electrodes.

Figure 2.9 XPS of BDD electrode after ‘standard activation protocol’

During EC activation, three mechanisms are believed to occur:

(i) oxidation of adsorbed organic molecule by OH*

(ii ) H-termination during cathodic pulsing19

(iii ) desorption of adsorbed organic compounds by electro-generated gaseous

species.

It was experimentally observed that the activation in basic solution has yielded better

reactivity when compared to acidic solutions. OH* formation occurs via the following

chemical reaction:22

H2O OH* + H+ + e-

The OH* generated is an oxidant and it oxidizes the adsorbed organic molecule to CO2.

OH* + R mCO2 + nH2O + e-

0 200 400 600 800 1000

0

1000

2000

3000

4000

5000

O 1s

Inte

nsity (

a. u

.)

Binding energy (eV)

C 1s

Electrochemical activation of diamond electrodes

Chapter II Page 44

The concentration of OH* increases progressively until a constant value is reached due to

mass transfer limitations or chemical destruction of the OH* formed. At higher pH more OH*

is generated. LiClO4 is a non-electroactive species and hence the only oxidant species

prensent is OH*.

The double layer capacitance of the electrode was measured before and after activation. Two

techniques were used for measurement of capacitance: CV (the electrode was scanned in 0.5

M LiClO4 solution at 100 mV s-1) and charge-discharge curves (I-t plot, where the electrode

was held at 0 V for 0.005 s and at 0.05 V for 0.005 s). Figure 2.10 shows the voltammogram

of ‘as grown’ and activated BDD electrodes. There was no significant difference in the

capacitance value (less than 0.5%) before and after activation. XPS analysis, high k0 value

and negligible change in double layer capacitance indicate that there is no significant

difference between ‘as grown’ electrode and EC activated electrode in terms of

electrochemical reactivity and surface termination.

Figure 2.10 Cyclic voltammogram of BDD electrode before and after ‘standard activation

protocol’ scanned at 100 mV.s-1.

Electrochemical activation of diamond electrodes

Chapter II Page 45

In order to investigate the action of EC activation on the electrode surface, a biofilm was

grown on the electrode surface. BDD electrodes were horizontally immersed in a domestic

aquarium (aerated and maintained at 23-28 °C) containing fish and aquatic plants for 90 days.

The electrodes were then rinsed in DI water, dried and were observed under optical

microscope. A biofilm had formed on the surface which consisted of several microbial

communities. SEM images (figure 2.11) shows that the BDD surface was masked by the

microbes.

(a)

(b)

(c)

(d)

Figure 2.11 Optical microscopic (x10 magnification) and SEM images of electrode deposited

with biofilm (a,c) and after three ‘standard activation protocols’ (b,d).

Electrochemical characterization such as EIS and CV were performed between various steps

of EC activation. Table 2.5 summarizes ∆Ep and a k0 values after each activation steps. It took

three ‘standard activation protocols’ to clean the electrode completely and get a ∆Ep value (in

[Fe(CN)6]3-/4- solution) of 60 mV and k0 > 0.1 cm.s-1. It was observed from SEM and optical

images that the microorganisms had been completely removed by the three of ‘standard

Electrochemical activation of diamond electrodes

Chapter II Page 46

activation protocol’. Three processes could explain the activation process: OH* might have

destroyed the microorganisms, electro-generated O2 and H2 gas bubbles helped in desorbing

the attached life and/or cathodic pulses have H-terminated the surface.

Table 2.5 Comparison of the peak-to-peak voltage separation (∆Ep) and electron transfer rate

(k0) values for an electrode deposited with biofilm and after each ‘standard activation

protocol’.

〉EP (mV) k0 (cm.s-1)

Biofilm 381 3.1 x 10-4

After Activation 1 178 1.02 x 10-3

After Activation 2 83 1.3 x 10-2

After Activation 3 60 0.11

The same approach was also conducted in urine. Four H-terminated BDD samples were

prepared: sample 1 is an ‘as grown’ BDD electrode, samples 2, 3 and 4 are electrodes fouled

by 5 CV scans in human urine.

Electrochemical activation of diamond electrodes

Chapter II Page 47

(a) (b)

(c)

(d)

Figure 2.12 SEM images of as grown electrode (a), electrode after being fouling in urine (b)

after activation in urine (c) and after activation in LiClO4 (d)

Samples 3 and 4 were cleaned using ‘standard cathodic protocol’ in urine and ‘standard

activation protocol’ respectively. Activation of electrode in urine is explained in detail in

section 2.6. The fouled sample (no. 2) appeared very cloudy under SEM (figure 2.12) and the

underlying BDD film could not clearly be observed suggesting the presence of nonconductive

biomolecule (enzyme, protein, fat etc.) layer over the BDD film.

Samples 1 and 4 exhibit similar aspects suggesting that sample 4 has been almost cleaned

with the help of the OH* generated that has oxidized the adsorbed biomolecules, the

desorption of the adsorbed species by electro-generated H2 and O2 bubbles and of the cathodic

pulse H-termination (observed by high k0 values). Sample 3 was almost cleaned with a less

cloudy appearance. There was no generation of OH* as the activation was cathodic. Hence

the cleaning was due to desorption of the adsorbed species by electro-generated H2 bubbles

and H-termination of surface (observed by high k0 values).

Electrochemical activation of diamond electrodes

Chapter II Page 48

2.6 In-situ activation in biological fluids

A set of 9 trials each comprising of 3 CV scans were performed in human urine and the

current densities of the 1st peak (P1) of the first scan were recorded. The electrode was then

activated in urine for all the trials except for trial no: 4, 7, and 9. The EIS of the electrode was

recorded in between each trials to enable the comparison of the k0 values between activated

and non-treated trials.

Activation in human urine was performed using a train of negative current pulses of 50

pulses. Each pulse had an amplitude of -20 mA.cm-2, a duration of 110 ms and a duty cycle of

90.91%. It was observed that the mean value of k0 for an activated electrode was 0.094 ± 006

cm.s-1. When the electrode was not activated this value was reduced by a factor of 40 with

respect to that of an activated electrode. From figure 2.13 it appears that once the electrode is

activated following a non-treated trial, one can bring back the initial reactivity of the electrode

as well as a k0 value close to mean values.

Figure 2.13 Electron transfer rate k0 measured after each trial. For trial no: 4, 7 and 9, the

electrode was not activated in urine. For all other trials, the electrode was activated in urine

and as a result the k0 of the electrode is close to 0.01 cm.s-1.

Similarly the mean current density J1 of the peak P1 for the activated trials was observed to

be 344.71 ± 9.7 µA.cm-2. For the non- treated trials, this value was only 3/4th of the activated

trials. This value was brought back to mean values by activation in urine as seen in figure

Electrochemical activation of diamond electrodes

Chapter II Page 49

2.14. Although cleaning in non-electroactive solution is preferable (as the peak values differ

by less than 0.01%), in-situ activation can be an alternative solution to cleaning owing to the

low standard deviation of J1.

Figure 2.14 Current density J1 (µA.cm-2 ) of the peak P1 measured from CV in urine versus

trial number. The peak values J1 of activated trails are close to 350 µA.cm-2 and if the

electrodes are not activated (trial no: 4, 7 and 9) this value dropped by 25%.

Another experiment used a BDD electrode of high reactivity (k0 = 0.1 cm.s-1) dipped in

bovine blood for two hours. This resulted in a drop of k0 to 0.007 cm.s-1. The EC activation

was then performed directly in bovine blood by applying a train of negative current pulses

consisting of 50 pulses where each pulse had an amplitude of -20 mA.cm-2, a duration of 110

ms and a duty cycle of 90.91% (same conditions as those for urine). EIS measurement showed

that the k0 value was brought back to 0.012 cm.s-1 after activation.

In summary, we concluded that this very novel activation process has very significant

potential in the monitoring of analytes in real samples such as blood, urine etc. because of (i)

the very short activation times and (ii) the tunable current density. However, in real samples,

adsorbents could accumulate on the surface of the electrode which can be desorbed by

applying stronger negative pulses (indicated by an increase in k0 value after activation).

Electrochemical activation of diamond electrodes

Chapter II Page 50

Conclusion

It has been demonstrated that this electrochemical treatment retrieves the lost reactivity of

BDD electrodes when aged in air as well as in solutions or fouled by a medium. The higher

the current density, the lower the time required to activate, and the better the result. By tuning

the above mentioned parameters (current density, pulse duration and number of pulses, and

type of the electrolyte, etc.), one can increase the charge transfer rate constant k0 to reach

values above 0.01 cm.s-1.

The other advantage of this technique is to enhance the reusability of the BDD electrode. As

opposed to other more conventional pre-treatments techniques reported in the literature, such

as anodic, cathodic or thermal ones, this novel electrochemical pre-treatment is relatively

simple, fast, and requires a minimum of energy.

The real breakthrough is that it can be successfully done directly in the measuring solution

especially in a biofluid. As a result this electrochemical activation can be performed prior to

analytical measurements to ensure reliable and reproducible results, especially when the

electrode has not been used for a long period of time. Although an aqueous electrolyte

containing non-electroactive species is preferred for EC activation, it can also be done in

biological fluids such as blood, urine etc, thereby opening the field for in-situ analysis. An

example of analyte quantification (uric acid concentration) with in-situ cleaning in human

urine is described in detail in chapter 4. It also opens the door to long term field

measurements where electrodes are prone to fouling when immersed for long periods of time

in liquids, including on-line measurement since electrodes may be regenerated very quickly

in-situ within biological medium.

Electrochemical activation of diamond electrodes

Chapter II Page 51

Bibliography

(1) Iniesta, J.; Michaud, P. A.; Panizza, M.; Comninellis, C. Electrochemistry Communications 2001, 3, 346–351.

(2) Montilla, F.; Michaud, P. A.; Morallon, E.; Vazquez, J. L.; Comninellis, C. Electrochimica Acta 2002, 47, 3509–3513.

(3) Iniesta, J.; Michaud, P. A.; Panizza, M.; Cerisola, G.; Aldaz, A.; Comninellis, C. Electrochimica Acta 2001, 46, 3573–3578.

(4) Davis, J.; Compton, R. G. Analytica Chimica Acta 2000, 404, 241–247.

(5) Hin, F. Y. Y.; Lowe, C. R. Analytical chemistry 1989, 59, 2111–2115.

(6) Yantasee, W.; Charnhattakorn, B.; Fryxell, G. E.; Lin, Y.; Timchalk, C.; Addleman, R. S. Analytica chimica acta 2008, 620, 55–63.

(7) Wang, J.; Hutchins, L. D. Analytical chemistry 1985, 57, 1536–1541.

(8) Vanhove, E.; de Sanoit, J.; Arnault, J. C.; Saada, S.; Mer, C.; Mailley, P.; Bergonzo, P.; Nesladek, M. Physica Status Solidi (a) 2007, 204, 2931–2939.

(9) Salazar-Banda, G. R.; Andrade, L. S.; Nascente, P. A. P.; Pizani, P. S.; Rocha-Filho, R. C.; Avaca, L. A. Electrochimica Acta 2006, 51, 4612–4619.

(10) Duo, I.; Levy-Clement, C.; Fujishima, a.; Comninellis, C. Journal of Applied Electrochemistry 2004, 34, 935–943.

(11) Ramesham, R. Thin Solid Films 1998, 315, 222–228.

(12) Popa, E.; Kubota, Y.; Tryk, D. A.; Fujishima, A. Analytical chemistry 2000, 72, 1724–1727.

(13) Salimi, A.; Mamkhezri, H.; Hallaj, R. Talanta 2006, 70, 823–32.

(14) Anderson, N. G.; Anderson, N. L.; Tollaksen, S. L. Clinical chemistry 1979, 25, 1199–210.

(15) Kraft, A. International Journal of Electrochemical Science 2 2007, 2, 355–385.

(16) Tryk, D. a.; Tsunozaki, K.; Rao, T. N.; Fujishima, a. Diamond and Related Materials 2001, 10, 1804–1809.

(17) Rodrigo, M. a.; Michaud, P. a.; Duo, I.; Panizza, M.; Cerisola, G.; Comninellis, C. Journal of The Electrochemical Society 2001, 148, D60.

(18) Mahé, E.; Devilliers, D.; Comninellis, C. Electrochimica Acta 2005, 50, 2263–2277.

Electrochemical activation of diamond electrodes

Chapter II Page 52

(19) Hoffmann, R.; Kriele, A.; Obloh, H.; Hees, J.; Wolfer, M.; Smirnov, W.; Yang, N.; Nebel, C. E. Applied Physics Letters 2010, 97, 052103.

(20) Panizza, M.; Cerisola, G. Electrochimica Acta 2005, 51, 191–199.

(21) Canizares, P.; Garcia-Gomez, J.; Lobato, J.; Rodrigo, M. A. Industrial & Engineering Chemistry Research 2003, 42, 956–962.

(22) Marselli, B.; Garcia-Gomez, J.; Michaud, P.-A.; Rodrigo, M. A.; Comninellis, C. Journal of The Electrochemical Society 2003, 150, D79–D83.

CHAPTER III

Microelectrode: Design, fabrication and

characterization

Microelectrode: Design, fabrication and characterization

Chapter III Page 54

Microelectrode: Design, fabrication and characterization

Chapter III Page 55

Introduction

Microelectrodes made from noble metals like platinum1, gold2, as well as silicon3 have a long

history as neural electrodes. However, they are known to suffer from long term stability.

Other materials such as functionalized hydrophilic Carbon nanotubes (CNT) have also been

used in microelectrode arrays (MEA) as a novel prototype neural interface due to their high

charge injection limit.4 However, the cytotoxicity of CNT and the biocompatibility issue,

especially during long term in-vitro measurements, remains an issue to address.5 Titanium

nitride and iridium oxide microelectrodes are extensively used for electrophysiological

applications owing to their enhanced charge injection limit.6 Although these materials possess

high double layer capacitance, the reduced potential window and high background current

limit their use in electro-analytical applications.

Boron-doped diamond exhibits superior electrochemical properties over other conventional

electrode materials, in particular its bio-inertness,7 corrosion resistance and long term

stability.7–9 Thus the advantage of using MEAs over macro-electrodes has been further

improved by combining their unique properties resulting from geometrical characteristics

with the excellent electrochemical properties of BDD materials.10–13 Several interconnected

and individually addressable diamond MEAs have been described in the literature: Hess et al.

have reported on the fabrication and characterization of a 10 channel diamond on polymer

MEA,14 Gao et al. on a 4 channel MEA to detect catecholamine.15 However for

electrophysiological applications, much more miniaturized electrodes are desirable to record

more accurately cell networks or organ (spinal cord, retina, etc.) communication mechanisms.

This led us to the fabrication of a BDD ultra-MEA (UMEA) consisting of an array of 64

multiple electrodes, where each electrode is individually connected to a single electrical

connector and then to a multichannel readout system. In additional to individually addressable

UMEAs, strip microelectrodes and ultramicroelectrodes and interconnected MEAs were

fabricated for electro-analytical applications.

The detailed fabrication processes of the different types of microelectrodes and MEAs are

described in this chapter. The fabricated electrodes were characterized using SEM and

electrochemical techniques such as cyclic voltammetry (CV) and electrochemical impedance

spectroscopy (EIS) to appreciate the electrodes’ performances namely their limiting currents

(i lim), electron transfer rates (k0), electrochemical windows and background currents (iBG) etc.

Microelectrode: Design, fabrication and characterization

Chapter III Page 56

They are the most efficient techniques to detect any cracks or discontinuity of the insulating

layer and analyzing EC properties of individual BDD UMEs.

3.1 Technological process

This work was carried out in close collaboration with ESIEE, Paris (École Supérieure

d'Ingénieurs en Électronique et Électrotechnique) where the clean room facilities as well as

know-how were used. The main goal was not to optimize the size of the electrode, but rather

to fabricate a working device with minimal leakage current and active electrodes. Several

types of designs were explored. Here we will describe just 2 types of designs (1 and 2) used to

fabricate the microelectrode strips and MEAs to highlight what are the difficulties in

developing such a process, and how alternatives have to be sought. In design 1 diamond was

grown after the deposition of the metal contacts and passivation layer, whereas in design 2,

diamond was grown prior to the deposition of metal tracks and passivation layer.

3.1.1 Design 1

Figure 3.1 shows the schematics of the fabrication process associated with design 1. Silicon

dioxide insulation layer was grown on 4 inch silicon wafers. The wafers were oxidized in an

oxidation oven at 1323 K, under a H2O vapor flow to form a 500 nm thick silicon dioxide

layer.16 The process time was 35 minutes.

Titanium/platinum/titanium (Ti/Pt/Ti) tracks were deposited by lift-off on the top of the oxide

layer. Clariant Nlof 2020 was used as the sacrificial photoresist material. It was observed that

diamond exhibits poor adhesion on Pt and SiO2 layers and hence Ti/Pt/Ti assembly was used.

The Ti layer also prevents the formation of Platinum silicide, which display a quite high

resistivity.17 The thickness of the Ti and Pt layers were 50 and 150 nm respectively and 10 -

20 µm wide. A 500 nm thick silicon dioxide layer was deposited onto the substrate in order to

isolate the metal tracks from the electrolyte solution. Then a conventional lithography process

was carried out to define the opening of the oxide layer (5-25 µm in diameter).

Detonation diamond nanoparticles were spread onto the processed wafers.18 Diamond

nanoparticles were dispersed in deionized water, and polyvinyl alcohol (PVA) was added to

the solution in adequate amount and dissolved in colloidal solution by sonication. A 4 inch

silicon wafer was seeded with 0.01% wt/wt diamond nanoparticles, 1.5% wt/wt PVA using

spin coater. The nanoparticles were fixed on the substrate by subjecting it to short diamond

Microelectrode: Design, fabrication and characterization

Chapter III Page 57

growth (5-10 minutes) in a SEKI 6500 reactor (later known as “SEKI”). The samples were

exposed to hydrogen (99.4%)/methane (0.6%) plasma. The typical density range of the

nanoparticles was of 108- 1011 cm-2. The microwave power was adjusted to 3000 W, gas flow

at 200 SCCM, pressure to 35 mBar and temperature to 650 °C. After the growth period, the

PVA is completely removed from the surface and nanoparticles had grown slightly and fixed

on silicon wafer.

Next, an aluminum hard mask was deposited over the electrode areas by photolithography and

the diamond nanoparticles outside these protected areas were etched away using ultra short

RIE under oxygen/argon plasma.19 The protective aluminium layer was removed by wet

etching. The diamond growth was resumed with the conditions described in chapter 1 section

1.3. The fabricated diamond electrodes exhibited a thickness of approximately 300 nm and

diameter varying from 10 – 100 µm.

Figure 3.1 Schematics of diamond microelectrode fabrication process (Design 1)

Figure 3.2 a shows the SEM image of an ultramicroelectrode and 3.2 b that of a

macroelectrode. In both cases diamond film was grown over titanium film. Non-continuous

Microelectrode: Design, fabrication and characterization

Chapter III Page 58

diamond layers with impurities and pinholes are visible on diamond layer in both macro and

micro-electrodes. Diamond deposition is influenced negatively as titanium shows an

interaction with the carbon of the gas phase.20 As a result diamond nucleation is delayed and

the adhesion of the formed layers is poor, which is partly due to the occurring deformation of

the substrates but also to the different thermal expansion coefficients of the different

materials. The delay in diamond nucleation might be due to TiC formation. Diamond can

nucleate only after saturation of surface and when TiC layer has attained certain thickness.

The TiC layer is rough and porous and tends to split off easily.

a b

Figure 3.2 SEM images of BDD films grown over titanium: (a) microelectrode, (b)

macroelectrode.

The electrochemical characterization techniques are described in detail in chapter 1 (section

1.4) and 3 (section 3.3). Cyclic voltammogram of the microelectrodes were recorded in 0.5 M

aqueous LiClO4 solution and all the electrodes exhibited an electrochemical window less than

3 V. Figure 3.3 a shows voltammogram of one microelectrode in a 64 channel MEA (20 µm

in diameter) at 200 mV.s-1. The reason for the reduced window (compared to 3.4 V of BDD)

might be due to pinholes in the diamond layer so that the titanium tracks were in contact with

the electrolyte (as seen in SEM images – figure 3.2 a). Also the double layer capacitance CD

calculated (4 nF) at this scan rate was few orders of magnitude higher than that of diamond. It

was observed that there was interconnection between different electrodes with a resistance as

low as few kilo ohms between 2 electrodes. The steady limiting current (in 1mM Fe(CN)64- in

0.5 M KCl solution at 100 mV.s-1) was observed to be 160 nA (against theoretical value of 2.6

nA).

Microelectrode: Design, fabrication and characterization

Chapter III Page 59

Figure 3.3 (a) Potential window observed by cyclic voltammetry (scan rate=200 mV.s-1) in

0.5 M aqueous LiClO4 solution and (b) Nyquist plot (experimental data and fitted data) of an

ultramicroelectrode of diameter 20 µm

Microelectrode: Design, fabrication and characterization

Chapter III Page 60

The electron transfer rate k0 and CD estimated from Nyquist plots (figure 3.3 b) were 0.005

cm.s-1 and 1.5 nF. These results suggest that there is an interconnection between different

electrodes and the interconnection might have arisen due to change in surface conductivity of

oxide layer in hydrogen/methane/TMB plasma.21 Pinholes in diamond film reduced

electrochemical window and electrode short circuit led us to redesign the complete

technological process. Diamond films were grown at more than 650 °C and this high

temperature might influence the physical and chemical properties of other layers (metallic and

dielectric). In design 2 diamond, being very robust material was grown first followed by other

layers.

3.1.2 Design 2

The process flow of design 2 is similar to design 1 (section 3.1.1). Silicon dioxide layer was

grown on 4 inch silicon wafers by wet oxidation. After diamond nanoparticle fixation, an

aluminum hard mask was deposited over the electrode areas by photolithography and the

diamond nanoparticles outside these protected areas were etched away using ultra short RIE

under oxygen plasma. The metal hard mask was then chemically removed to reveal the

diamond nanoparticles patterns, on which diamond electrodes were grown in the SEKI

diamond growth reactor. The fabricated diamond electrodes, typically, exhibit a thickness of

approximately 300 nm over 200 micrometers in diameter. The electrodes were then

individually contacted from the top by the deposition of Ti (50 nm) /Pt (150 nm) metal tracks

using the lift-off process and Clariant Nlof 2020 as photoresist material. Contact on the

electrodes was achieved by deposition of a metal ring across the edges of the electrodes.

Finally a 600 nm thick silicon nitride (Si3N4) layer was deposited by CVD onto the substrate

in order to isolate the metal tracks from the electrolyte solution. Several insulating passivation

layers were tested including photo-resist SU-8, silicon oxide etc., but leakage capacitive

and/or faradic current were observed. Si3N4, having almost twice the dielectric constant of

SiO2, was chosen as the passivation layer because of lower leakage current and high

resistance to oxidation.22 Openings of the contact areas and of the diamond electrodes were

achieved using local etching of the Si3N4 layer by RIE with SF6 gas. It is this process step that

determined the final diameter of each individual diamond electrode. Finally the photoresist

used to selectively etch the nitride layer was removed and the UMEA was diced using a

diamond saw. The detailed fabrication process according to design 2 is depicted in Figure 3.4.

The samples were then cleaned in piranha solution (mixture of 3:1 concentrated sulfuric acid

Microelectrode: Design, fabrication and characterization

Chapter III Page 61

to 30% hydrogen peroxide) for 10 minutes and was exposed to hydrogen plasma to hydrogen

terminate the BDD films.

Several UMEs (strip electrodes and multi-electrode arrays) with diameters varying from 2 µm

to 100 µm have been fabricated using this technique. Most of these electrodes exhibited

typical BDD responses. The complete characterization of one such 8x8 diamond UMEAs

with diamond electrodes of 14µm in diameter and 100 µm of inter-electrode pitch is discussed

in detail in section 3.2 and 3.3.

Figure 3.4 Schematics of diamond microelectrode fabrication process (design 2)

3.2 SEM Characterization

Figure 3.5 a shows the SEM image of 8x8 UMEA along with the tracks, where one can

observe that each electrode is well separated (100µm pitch).The diamond crystals are highly

faceted with an average grain size of 100 nm as seen in Figure 3.5 d at higher magnification.

Here neither major cracks nor pin-holes were visible, under SEM or optical microscopy, nor

Microelectrode: Design, fabrication and characterization

Chapter III Page 62

on the tracks, nor on the passivation layer nor on the electrode surface as seen in Figures 3.5 b

and 3.5 c. The electrode surface appears bright under SEM owing to the fact that the surface is

conductive in steady state with respect to the outer passivation. In fact the observed surface

conductivity may be associated to the hydrogen (H) termination of the electrode surface just

after growth. A negative electron affinity is generated on the H terminated surface which

increases the density of the back scattered electron.23 The electron affinity of an H terminated

diamond surface can go up to -1.3 eV.23,24 When compared to design 1, we get a

polycrystalline film.

(a) (b)

(c) (d)

Figure 3.5 SEM image of an 8x8 UMEA along with the tracks (a) and magnified SEM image

of a single electrode (b, c and d).

Microelectrode: Design, fabrication and characterization

Chapter III Page 63

3.3 Electrochemical characterization

We have used electrochemical (EC) characterization techniques such as cyclic voltammetry

(CV) and electrochemical impedance spectroscopy (EIS) to appreciate the electrodes

performances namely their limiting currents (ilim), electron transfer rates (k0), electrochemical

windows and background currents (iBG) etc. They are the most efficient techniques to detect

any cracks, passivation films, surface contamination or discontinuity of the insulating layer

and analyzing EC properties of individual BDD UMEs.

3.3.1 Background current

CV was recorded on all the electrodes in 0.5 M aqueous LiClO4 solution to ascertain the

accessible EC window. Most of the electrodes exhibit a typical BDD window of over 3 V (in

aqueous solution). CV of one such electrode is shown in Figure 3.6 where the electrode was

scanned at 200 mV.s-1(details of experimental procedures are described in chapter 1 section

1.4.2).

Figure 3.6 Typical potential window observed by cyclic voltammetry (scan rate=200

mV.s-1) in LiClO4 solution where i is the current in nA and E is the voltage in volts

versus an Ag/AgCl reference electrode.

Microelectrode: Design, fabrication and characterization

Chapter III Page 64

One unexpected characteristic appeared on this first prototype, as few electrodes exhibited a

reduced potential window with a value typically equal to that of platinum electrodes. For this

reason it was chosen not to assess further those electrodes and they hence appear as white

spots on figures (3.7, 3.9 and 3.12), although of course they could be fully characterized but

not relevant with the properties of diamond. This was associated with cracks or pinholes in

the passivation layer so that the platinum tracks were in contact with the electrolyte. This,

although, not observed with imaging techniques becomes easily identified with EC

characterization. Other technique to detect the cracks or pinholes is fluorescence confocal

laser scanning microscopy as demonstrated by Rudd et al. on platinum UMEA.25

Figure 3.7 RGB model of an 8x8 electrode array where each electrode is represented by

a spot and the red component value corresponds to their respective iBG value. (In this

first prototype, and although measurable, white spots correspond to electrodes that were

not exhibiting the EC window of diamond, as attributed to Pt shorts from leaky tracks,

thus not relevant for the comparison).

Microelectrode: Design, fabrication and characterization

Chapter III Page 65

During CV characterization of each electrode, a transient current flows within the potential

window when the potential is varied, as ions move to the surface forming a double layer.26

The charging and discharging of this double layer constitutes the background current within

the potential window. The background current iBG was calculated for the electrodes of the

array and was observed to be 94±65 pA at a scan rate of 200 mV.s-1. iBG of the 8x8 matrix is

depicted in figure 3.7. RGB color model was used to indicate the amplitude of iBG of each

electrode. The green and blue color component has a value = 0 and red component vary from

0 to 255 corresponding to the amplitude of iBG. The darker the spot, the lower the background

current.

3.3.2 Steady state limiting current

Ideally, the limiting current of a UME is independent of the scan rate.27,28 The radial diffusion

behavior of the UME enhances the mass transfer thus promoting steady-state electrochemical

comportment. Hence, these microelectrodes exhibit typical sigmoidal shape in CV with a

small hysteresis coming from double layer capacitance. On the other hand a macro-electrode

shows a peak and the value of the peak current is directly proportional to the square root of

the scan rate as the planar diffusion behavior limits the current.

Figure 3.8 Cylcic voltammogram of an electrode at 25, 50 and 100 mV.s-1 in 0.5 M KCl

aqueous solution containing 1 mM Fe(CN)64- ion.

Microelectrode: Design, fabrication and characterization

Chapter III Page 66

Steady state voltammograms were recorded using CV in 1mM Fe(CN)64- in 0.5 M KCl

solution at 100 mV.s-1. The limiting current ilim of the 8x8 matrix was recorded and the

average limiting current value was of 1.83±0.2 nA as opposed to theoretical limiting current

of 1.8 nA (Diffusion coefficient of ferrocyanide = 6.67 x 10-6 cm² s-1 29). Furthermore they

varied from electrode to electrode with a standard deviation of 190 pA (Table 3.1).

Figure 3.9 RGB model of an 8x8 electrode array where each electrode is represented by

a circle and the green component value is inversely proportional to respective ilim value.

Figure 3.8 shows CV of an electrode at various scan rates. When the diffusion layer d is

greater than the electrode radius r, radial diffusion predominates and if r >> d, planar

diffusion predominates, where ‘t’ is the duration of experiment (refer to chapter 1 section

1.6). At slower scan rates of 25, 50 and 100 mV.s-1 the diffusion layer thickness d >> radius r

of the electrode and hence radial diffusion predominates and as a result a sigmoidal

voltammogram is observed. ilim observed was 1.6 nA and 1.66 nA at 25 mV.s-1 and 100 mV.s-

Microelectrode: Design, fabrication and characterization

Chapter III Page 67

1, respectively . Although a slight deviation from the ideal scan rate independent behavior was

observed, the current density variation is negligible (below 4%) when compared to a macro-

electrode at these scan rates.

Figure 3.9 represents an RGB color model indicating the value of ilim of each electrode where

red and blue color component has a value = 0 and green component vary from 0 to 255 and is

inversely proportional to the amplitude of ilim.

3.3.3 Electrochemical impedance spectroscopy

The Nyquist plots obtained from EIS of the UMEA were used to calculate the k0 of each

electrode. Electrochemical interfaces can be classically modeled by the Randles equivalent

electric circuit. The latter consists of 4 electronic components: RS the electrolytic resistance,

RT the charge transfer resistance, CD the double layer capacitance and ZW the Warburg

element. The diameter of the semi-circle portion corresponds to the transfer resistance.30,31

Figure 3.10 Nyquist plot (experimental data and fitted data) of an ultramicroelectrode.

Microelectrode: Design, fabrication and characterization

Chapter III Page 68

From the Nyquist plot of all those electrodes, as exemplified in figure 3.10, it was observed

that the straight line (Warburg element) almost disappeared since the linear correlation of Re

(Z) vs -img (Z) corresponds to a linear semi-infinite diffusion process (transient mass

transfer).32 Elsewhere, the semi-circular shape of the spectrum is not depressed thus

highlighting the pure capacitive behavior of the interface. The rate of mass transport to and

from the electrode is greater in UMEs when compared to macro-electrodes. Hence the

Randles circuit of a UME gets modified to a 3 impedance component (RS, RT and CD, the

Warburg diffusion impedance being deleted– refer figure 3.11).

An R(CR) model circuit was used to fit the experimental curves and the impedance value

obtained for RS, RT and CD were 414 ohm, 14.7 Mohm and 132 pF respectively. The ぬ2 error

was suitably low (ぬ2 < 10−4), and the error associated with each element was below 5%. The

electron transfer rate of this electrode was calculated to be 0.012 cm s-1. The semicircle

impedance spectra could also favorably be used for bio-sensing applications where one can

observe the change in the k0 value.30 The electrode electron transfer rate k0 is calculated using

equation (1.1) and variation of k0 along the electrodes in the chip is shown in Figure 3.12. The

average value obtained for k0 is of 0.013 cm s-1 which is close to the values reported by other

groups for an H terminated BDD electrode .33–35 The mean values and standard deviations of

iBG, ilim, and k0 are summarized in table 3.1.

Figure 3.11 3 Component of Randles’s circuit of ultramicroelectrode where RS is the

electrolytic resistance, RT is the charge transfer resistance and CD is the double layer

capacitance.

Microelectrode: Design, fabrication and characterization

Chapter III Page 69

k0 / cm s-1

1 2 3 4 5 6 7 8

A

B

C

D

E

F

G

H

0.0045

0.0146

0.0248

0.0349

0.045

Figure 3.12 RGB model of an 8x8 electrode array where each electrode is represented

by a circle and the blue component value corresponds to their respective k0 value.

Table 3.1 Mean value and standard deviation of the background current iBG, limiting current

i lim, and transfer rate k0.

Parameters Mean value Standard deviation

iBG 94.33 pA 64.85 pA

i lim 1.83 nA 0.19 nA

k0 0.0132 cm.s-1 0.008 cm.s-1

The electron transfer rate of the electrodes was further enhanced by applying the

electrochemical treatment mentioned in chapter 2. A train of 50 current pulses of alternating

amplitude (10 mA.cm-2 and -10 mA.cm-2) and of equal duration (100 ms) was applied

between each working microelectrode and the counter electrode in 0.5 M aqueous LiClO4

solution (standard activation protocol). After this treatment the k0 values of the electrodes

have increased by several folds and ilim measured was close to theoretical values (1.8 nA).

The EC treatment cleans the electrode surface further and hence improves the reactivity of the

electrodes.

Microelectrode: Design, fabrication and characterization

Chapter III Page 70

It has been observed that the values of k0, ilim, CD and iBG were somehow related to each other

for different electrodes. The variation of these values from electrode to electrode cannot be

explained by conventional EC and SEM characterization. These variations might be due to

difference in intrinsic conductivity, crystalline orientation and/or charge transfer constant of

individual grains. This could be associated with varying boron intake, even at high boron

concentrations, leading to varying electrical and electrochemical characteristics of diamond

grains. So, comparing the grain and the microelectrode sizes, conversely to macro-electrodes,

this grain distribution may statically markedly vary between each electrode and thus strongly

affect their electrochemical behavior.

A comparison is made to correlate different parameters such as the k0, ilim, CD and iBG of 2

electrodes and is summarized in table 3.2. The k0 and CD values were obtained from Nyquist

plots.

The background current (which is attributed by the double layer capacitance) is higher when

CD is higher since capacitive current represent the main background contribution owing to the

weak charge transfer associated to media hydrolysis. However, the difference in the values of

double layer capacitance measured using EIS and CV is 49% and 21% for electrode 1 and 2

respectively. This difference is coherent, since CV measurements are less sensitive to the

capacitive behavior (which is dominated by the faradic response) and just allow estimating

the order of magnitude of the capacitance.

From the limit current recording from CV curves, one can estimate the effective surface area

of the electrode, assuming a purely Nernstian electrochemical behavior (large charge transfer

rate). Typically, this effective surface area depends up on the roughness of the electrode but

also on the ability of the electrode toward charge transfer. Indeed, limiting current and

effective surface (due to their proportional relationship) vary much less than the measured k0.

Such a divergence can be explained through one fundamental aspects: first the limiting

current is measured in the diffusion limitation part of the CV (e.g. the plateau, at high over

potential) whereas k0 is calculated from the EIS spectrum obtained at the half-wave potential

of the redox couple under a charge limitation behavior. Thereby, such differences in the

variations of the limiting current and on the k0 clearly highlight the heterogeneities of the

charge transfer resistance over the electrode surface and thus the accentuated and discretized

role of the grain distribution on microelectrode behavior.

Microelectrode: Design, fabrication and characterization

Chapter III Page 71

Table 3.2 Comparison of different parameters such as the electron transfer rate k0, limiting

current ilim, double layer capacitance CD and background current iBG of 2 electrodes.

Parameters Electrode

1

Electrode 2 Difference %

x/x1

k0 (cm s-1) 0.012 0.017 42

i lim (nA) 1.74 1.76 11.5

CD (pF) 132 197 49

iBG at 200 mV.s-1

(pA)

63 76 21

3.4 All diamond microelectrode arrays

One of the disadvantages of individual microelectrodes is relative reduced currents they can

drive (pico to nanoamperes), due to their small surface area.36 Hence a noise canceller or

signal amplifier has to be included to enhance the quality of the electrochemical recordings.

To overcome this problem we have designed interconnected BDD microelectrode arrays

using intrinsic diamond as the passivation layer. In theory, the microelectrode arrays behave

as an independent microelectrode as their diffusional fields do not overlap at suitable scan

rate.11 In an interconnected MEA, the overall output of the current is the sum of the steady-

state redox currents on individual electrodes of the array.

Figure 3.13 summarizes the complete fabrication process. The technological processes are

similar to design 1 and 2. Silicon wafers were nanoseeded with diamond nanoparticles and

BDD fil ms were grown in the SEKI MPECVD diamond growth reactor with TMB gas (refer

to chapter 1 section 1.3). The BDD film has a thickness of approximately 300 nm. Chromium

mask was deposited on BDD layer with Si3N4 as the sacrificial layer. Intrinsic diamond was

grown on top of this layer (growth conditions were similar to that of boron doped diamond

except from the absence of TMB doping gas, in a twin SEKI 6500 4 inch reactor –“THOR”

that was never exposed to doping gases). The intrinsic diamond film has a thickness of 800

nm. Finally the chromium mask was etched to expose the conductive BDD film. The samples

were then cleaned in piranha solution for 10 minutes and exposed to hydrogen plasma to

hydrogen terminate the BDD films. Each chip had a total of 58 microelectrodes of 17 µm in

diameter. Hence the total surface area of BDD film in contact with the electrolyte is 0.000132

cm². The inter-electrode distance is 60 µm. Figure 3.14 shows the SEM images of the MEA.

Microelectrode: Design, fabrication and characterization

Chapter III Page 72

The darker area (figure 3.14 a) is the intrinsic diamond and the lighter area is the boron doped

diamond. One can observe that there is clear-cut boundary between the doped and intrinsic

diamond which shows that there is no boron diffusion across the interface.

Figure 3.13 Schematics of interconnected diamond microelectrode array fabrication process.

a b

Figure 3.14 SEM images of MEA and individual BDD conducting electrode at 17 µm

in diameter

Microelectrode: Design, fabrication and characterization

Chapter III Page 73

The fabricated electrodes were characterized by EIS and CV to evaluate their electrochemical

properties (refer to section 3.3 for more details). The CV in 0.5 M aqueous LiClO4 solution

demonstrated a large potential window greater than 3 V (figure 3.15) and a double layer

capacitance of 345 nF (2.62 mF.cm-2) measured at scan rate of 100 mV.s-1. The electron

transfer rate k0 and CD estimated from Nyquist plot were 0.56 cm.s-1 and 227 nF (1.73 mF.cm-

2). Figure 3.16 shows the CV of MEA in 1 mM Fe(CN)63-/4- in 0.5 M aqueous KCl solution at

different scan rates: 10, 25, 50, 100, 200, 500, 1000 and 2000 mV.s-1. The shape of the CV of

the MEA depends on two parameters: thickness of the diffusion layer (depends on the scan

rate and potential window) and the inter-electrode distance (that has to be chosen to avoid any

coverage of the hemispherical diffusion profiles arising from each microelectrode). It is

observed from figure 3.16 that for most of the scans (scan rate < 500 mV.s-1) exhibited an

electrochemical behavior similar to that of a macro-electrode. The diffusion layer of each

microelectrode converges over each other to form a continuous diffusion layer. At higher scan

rates, the CV is similar to that of steady-state voltammograms. The steady state limiting

current of the MEA being equal to the sum of the limiting currents of the entire individual

electrode, it should in theory reach 127 nA (equation 1.9). The limiting current extracted for

Fe(CN)64- oxidation at 500 mV.s-1 from figure 3.16 is 22 µA. The peak current ip for

Fe(CN)64- oxidation at 10 mV.s-1 can be calculated from the following equation of a

macroelectrode is 9.1 nA:

ip= 2.69x105n3/2ACD1/2v1/2 (2.1)

Where D0 is the diffusion coefficient, C0 the concentration, and n the number of electrons

transferred per molecule of reactant, A the total surface area and v the scan rate. The value of

the peak current observed from figure 3.16 at this scan rate is 12.8 µA.

For an electrode of 17 µm in diameter the diffusion layer thickness (equation 1.5) d is 26.33

µm at 25 mV.s-1 scan rate and the MEA should exhibit sigmoidal voltammogram behavior

from this scan rate onwards. The high double layer capacitance, electron transfer rate and

steady state limiting current suggest that the electron transfer is not only the result of the 58

electrodes but also from the intrinsic diamond which serves as the passivation layer. As the

MEA is fabricated under hydrogen plasma, the H-terminated intrinsic (undoped) diamond

when immersed in redox electrolyte solutions shows reversible insulator - metal transitions37.

The hydrogen induced surface conductivity is due to a doping mechanism where the surface

adsorbates act as acceptors.38 The surface acceptors can be solvated ionic species that undergo

Microelectrode: Design, fabrication and characterization

Chapter III Page 74

electrochemical reactions after accepting an electron from the diamond valence band. This

increases of the overall active surface area could be attributed to the high electron transfer rate

and limiting current. This problem can be eliminated by selective O-termination of the

intrinsic layer.

Figure 3.15 Cyclic voltammograms of boron doped diamond MEA in 0.5 M aqueous

LiClO4 solution.

Figure 3.16 Cyclic voltammograms of boron doped diamond MEA in 0.5 M aqueous

KCl solution containing 1 mM Fe(CN)6 3-/4- at scan rate 10 mV.s-1 to 2 V.s-1.

Microelectrode: Design, fabrication and characterization

Chapter III Page 75

Conclusions

A novel, high yield and reproducible lithographic process has been used to fabricate the

UMEAs. BDD UMEAs suitable for use in electrochemical sensors were prepared by micro-

fabrication technique compatible with standard clean room technology. Topographical

characterization and detailed electrochemical study of individual UMEs revealed only few

faulty electrodes within an array. In electrochemical tests, the UMEs exhibited low

background currents, almost theoretical steady state limiting currents and fast electron transfer

rates (close to 0.01 cm s-1). Further improvement in these two values was achieved through

EC activation. The long term goal of this work was ultimately to develop biosensing

platforms for the monitoring of neural activities for electrophysiology. Electro-analytical and

electrophysiological applications of those microelectrodes and MEA platforms are mentioned

in detail in chapter 4 and 5.

Microelectrode: Design, fabrication and characterization

Chapter III Page 76

Bibliography

(1) Thiébaud, P.; Beuret, C.; Koudelka-Hep, M.; Bove, M.; Martinoia, S.; Grattarola, M.; Jahnsen, H.; Rebaudo, R.; Balestrino, M.; Zimmer, J.; Dupont, Y. Biosensors & bioelectronics 1999, 14, 61–65.

(2) Liu, S.; Chen, Y.; Fang, F.; Xu, J.; Sheng, G.; Yu, H.; Liu, G.; Tian, Y. Environmental Science & Technology 2009, 43, 1160–1165.

(3) Biran, R.; Martin, D. C.; Tresco, P. A. Experimental neurology 2005, 195, 115–126.

(4) Wang, K.; Fishman, H. A.; Dai, H.; Harris, J. S. Nano letters 2006, 6, 2043–20488.

(5) Abarrategi, A.; Gutiérrez, M. C.; Moreno-Vicente, C.; Hortigüela, M. J.; Ramos, V.; López-Lacomba, J. L.; Ferrer, M. L.; del Monte, F. Biomaterials 2008, 29, 94–102.

(6) Weiland, J. D.; Anderson, D. J.; Humayun, M. S. IEEE transactions on bio-medical engineering 2002, 49, 1574–9.

(7) Panizza, M.; Cerisola, G. Electrochimica Acta 2005, 51, 191–199.

(8) Yano, T.; Tryk, D. A.; Hashimoto, K.; Fujishima, A. Journal of The Electrochemical Society 1998, 145, 1870–1876.

(9) Swain, G. M. Analytical chemistry 1993, 65, 345–351.

(10) Soh, K. L.; Kang, W. P.; Davidson, J. L.; Basu, S.; Wong, Y. M.; Cliffel, D. E.; Bonds, A. B.; Swain, G. M. Diamond and Related Materials 2004, 13, 2009–2015.

(11) Lawrence, N. S.; Pagels, M.; Meredith, A.; Jones, T. G. J.; Hall, C. E.; Pickles, C. S. J.; Godfried, H. P.; Banks, C. E.; Compton, R. G.; Jiang, L. Talanta 2006, 69, 829–834.

(12) Tian, R.; Zhi, J. Electrochemistry Communications 2007, 9, 1120–1126.

(13) Pagels, M.; Hall, C. E.; Lawrence, N. S.; Meredith, A.; Jones, T. G. J.; Godfried, H. P.; Pickles, C. S. J.; Wilman, J.; Banks, C. E.; Compton, R. G.; Jiang, L. Analytical chemistry 2005, 77, 3705–8.

(14) Hess, A. E.; Sabens, D. M.; Martin, H. B.; Zorman, C. A.; Member, S. jounral of microelectromechanical systems 2011, 20, 867–875.

(15) Gao, Z.; Carabelli, V.; Carbone, E.; Colombo, E.; Dipalo, M.; Manfredotti, C.; Pasquarelli, A.; Feneberg, M.; Thonke, K.; Vittone, E.; Kohn, E. Journal of Micro-Nano Mechatronics 2011, 6, 33–37.

(16) Bongrain, A.; Scorsone, E.; Rousseau, L.; Lissorgues, G.; Gesset, C.; Saada, S.; Bergonzo, P. Journal of Micromechanics and Microengineering 2009, 19, 074015.

Microelectrode: Design, fabrication and characterization

Chapter III Page 77

(17) Sreenivas, K.; Reaney, I.; Maeder, T.; Setter, N.; Jagadish, C.; Elliman, R. G. Journal of Applied Physics 1994, 75, 232.

(18) Scorsone, E.; Saada, S.; Arnault, J. C.; Bergonzo, P. Journal of Applied Physics 2009, 106, 014908.

(19) Bongrain, A.; Scorsone, E.; Rousseau, L.; Lissorgues, G.; Gesset, C.; Saada, S.; Bergonzo, P. Journal of Micromechanics and Microengineering 2009, 19, 074015.

(20) Gerger, I.; Haubner, R.; Kronberger, H.; Fafilek, G. Diamond and Related Materials 2004, 13, 1062–1069.

(21) Vande Voorde, P.; Cox, M.; Greene, W. IEEE Electron Device Letters 1998, 19, 291–293.

(22) Ma, T. P. ieee transactions on electron devices 1998, 45, 680–690.

(23) Nebel, C. E.; Shin, D.; Takeuchi, D.; Yamamoto, T.; Watanabe, H.; Nakamura, T. Langmuir 2006, 22, 5645–5653.

(24) Maier, F.; Riedel, M.; Mantel, B.; Ristein, J.; Ley, L. Physical review letters 2000, 85, 3472–3475.

(25) Rudd, N. C.; Cannan, S.; Bitziou, E.; Ciani, I.; Whitworth, A. L.; Unwin, P. R. Analytical chemistry 2005, 77, 6205–6217.

(26) Qu, D.; Shi, H. Journal of Power Sources 1998, 74, 99–107.

(27) Colombo, E.; Men, Y.; Scharpf, J.; Pietzka, C.; Dipalo, M.; Herfurth, P.; Gao, Z.; Schneider, M.; Carabelli, V.; Carbone, E.; Kohn, E.; Pasquarelli, A. Diamond and Related Materials 2011, 20, 793–797.

(28) Caudlll, W. L.; Howell, J. O.; Wlghtman, R. M. Analytical chemistry 1982, 2532–2535.

(29) Konopka, S. J.; Mcduffie, B. Analytical chemistry 1970, 42, 1741–1746.

(30) Yun, Y.; Bange, A.; Heineman, W. R.; Halsall, H. B.; Shanov, V. N.; Dong, Z.; Pixley, S.; Behbehani, M.; Jazieh, A.; Tu, Y.; Wong, D. K. Y.; Battacharya, A.; Schulz, M. J. Sensors and Actuators B: Chemical 2007, 123, 177–182.

(31) Dumitrescu, I.; Unwin, P. R.; Macpherson, J. V. Electrochemistry Communications 2009, 11, 2081–2084.

(32) Siddiqui, S.; Arumugam, P. U.; Chen, H.; Li, J.; Meyyappan, M. ACS nano 2010, 4, 955–961.

(33) Fischer, A. E.; Show, Y.; Swain, G. M. Analytical chemistry 2004, 76, 2553–2560.

Microelectrode: Design, fabrication and characterization

Chapter III Page 78

(34) Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, S.; Show, Y.; Swain, G. M. Diamond and Related Materials 2003, 12, 1940–1949.

(35) Granger, M. C.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.; Koppang, M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M. Anal. Chem. 2000, 72, 3793–3804.

(36) Hayashi, K.; Takahashi, J.; Horiuchi, T.; Iwasaki, Y.; Haga, T. Journal of The Electrochemical Society 2008, 155, J240–J243.

(37) Shin, D.; Watanabe, H.; Nebel, C. E. Journal of the American Chemical Society 2005, 127, 11236–7.

(38) Ristein, J. Journal of Physics D: Applied Physics 2006, 39, R71–R81.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 79

CHAPTER IV

Diamond microelectrodes: Electroanalytical

application

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 80

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 81

Introduction

This chapter deals with an electro-analytical application of microelectrodes fabricated using

the technology described in chapter 3: Quasi real time quantification of uric acid in human

urine. Uric acid (UA) is the principal breakdown product of purine metabolism1 and the

normal UA concentration in human urine is around 2 mM.2 Overconsumption of protein,

overdrinking of alcoholic beverages, heavy stress and lack of exercise increases the amount of

uric acid in blood serum.3 Hyperuricemia, associated with renal disease, can cause gout,4

cardiovascular diseases,5 kidney stones6 etc. whereas hypouricemia can be due to Fanconi

syndrome,7 Nephritis8 and other kidney disorders. Hence it is important to monitor UA levels

in bodily fluids such as blood and urine.

Nowadays, the determination of uric acid in urine is performed in medical laboratories using

mainly spectrophotometric analysis methods. Other techniques reported in the literature

include chromatographic methods, capillary electrophoresis and electrochemical (EC)

methods.3 Enzyme-based quantification techniques are associated to oxidation of uric acid in

the presence of uricase yielding allantoin, carbon dioxide and hydrogen peroxide. However

there is also an interest in continuous monitoring of UA in urine, in particular for patients

admitted in intensive care units (ICU), where the early diagnostic of acute renal failure (ARF)

observed in up to 25% of the patients and it can have a major impact on the survival rate of

those patients.9 In this context electrochemical detection techniques are seen as a promising

alternative to conventional optical methods due to their good sensitivity, fast measuring time,

portability, low power consumption and cost effectiveness, thus enabling direct bedside

monitoring.

Various electrochemical approaches such as a polymer modified electrode,10–12 a chemically

modified electrode,13–16 an enzyme modified electrode,17 an electrochemical pre-treatment18

have been developed to detect UA. However UA coexists with ascorbic acid (AA) in

biological fluids and has got nearly the same oxidation potential.11 Although modified

electrodes show good selectivity to UA, complications such as adsorption, fouling etc. are

associated with those techniques.14 Simultaneous determination of dopamine, AA and UA

were also investigated using fast cyclic voltammetry (CV) and differential pulse voltammetry

(DPV) without any electrode modification.19 However, using this techniques, the analyte must

be diluted to prevent the surface from fouling.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 82

Popa et al. has achieved UA and AA peak separation by anodizing the diamond surface.2

However, at high pH values the peak separation was diminished thus making UA

quantification difficult in the presence of AA. Moreover, the analyte has to be diluted by

several thousand fold to obtain reliable results. Fast CV demonstrates the production of a very

reducible electro-active species that has been produced as a result of oxidation of UA.20

The typical CV response of UA shows one broad irreversible peak at slow potential sweeping

(down to 1 V.s-1) on classical macro-electrodes. By using fast CV (1 V.s-1), Dryhurst

demonstrated the existence of a weak reduction peak resulting from UA oxidation.20 Indeed,

the oxidation of C4=C5 bond of UA gives readily reducible bis-imine (on C4 atom) that may

undergo further irreversible chemical hydration reaction if not quickly electrochemically

reduced (figure 4.3). This bis-imine compound is highly reactive and readily reducible.

Complete hydration of bis-imine gives rise to uric acid-diol. Conversely, electrochemical

oxidation of AA is known to be highly irreversible.21 Hence fast CV may be used to

selectively determine the concentration of UA in the presence of AA.

The BDD microelectrode (of 40 µm diameter) strips fabricated using our technology (refer to

chapter 3) were used for the selective and sensitive detection of UA in the presence of low

and high quantities of AA. The low double layer capacitance of diamond reduces the

background current and potentially increases the signal-to-background ratio. Microelectrodes

show a decreased ohmic drop, a hemispherical diffusion layer and a fast establishment of a

steady-state signal when compared to macro-electrodes.22 This chapter discusses the

electrochemical characterization of the microelectrode, quantification of UA in presence of

low and high concentration of AA and in-situ cleaning techniques.

4.1 Electrochemical characterization of BDD microelectrode

Electrochemical characterization where carried out in order to know the reactivity of the

electrode, background current and the accessible electrochemical window. A low background

current, high reactivity and theoretical steady state limiting current (in a redox couple) is

essential for accurate and reproducible measurements.

Figure 4.1a shows the CV in 0.5 M aqueous LiClO4 solution demonstrating that the accessible

potential window of the BDD film is about 3.4 V with a background current of 30 pA at 0.2

V.s-1. The equivalent Randles circuit of this microelectrode is a three impedance component

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 83

(RS, RT and CD), where RS is the series resistance, RT the transfer resistance and CD the double

layer capacitance.

Figure 4.1 (a) Cyclic voltammogram of 40 µm diameter microelectrode in 0.5 M aqueous

LiClO4 solution at 0.2 V.s-1 and (b) experimental and fitted Nyquist plot obtained at open

circuit potential for a modulation of 10 mV rms in aqueous 0.5 M KCl solution containing

equimolar ferri/ferrocyanyde (1 mM).

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 84

An R(C/R) model circuit was used to fit the experimental curves and the impedance values

obtained for RS, RT and CD were 263 ohm, 1.4 Mohm and 183 pF respectively. The ぬ2 error

was suitably low (ぬ2鳥 鳥< 鳥10−4), and the error associated with each element was less than 5%.

Figure 4.1b shows the experimental and fitted Nyquist plot where the electrodes exhibit a very

fast electron transfer rate (k0) up to 0.02 cm.s-1. The steady state limiting current ilim is 4.7 nA

in a 0.5 M potassium chloride aqueous solution containing 1 mM Fe(CN)64- ion. The variation

of calculated ilim from theoretical ilim (5.1 nA ) for this electrode was, however, observed to

be less than 8% (Diffusion coefficient of ferrocyanide = 6.67 × 10−6 cm².s−1 23).

The potential window corresponds to that of diamond and low background current

demonstrates that neither cracks nor pinholes where detected in the passivation layer. High

electrochemical reactivity characterized by the quick electron transfer rate of 0.02 cm.s-1

makes this electrode sensitive to amperometric sensing of biomolecules such as uric acid and

ascorbic acid. Although several electrodes were tested, this particular electrode (40 µm in

diameter) was used for the electro-analytical based on its closeness to the theoretical limiting

current values and low background current.

4.2 Cyclic voltammogram of uric acid and ascorbic acid

UA and AA coexist in physiological liquid with overlapping oxidation potential on most

electrodes. We did some studies on diamond electrodes to analyze the overlapping CV

behavior.

UA and AA solutions were prepared daily by dissolving in phosphate buffered saline (PBS)

aqueous solution of pH 7.2. The electrode was scanned from -0.3 V to 1.4 V vs Ag/AgCl.

Figure 4.2 shows the CV of AA and UA recorded at low scan rate (0.1 V.s-1) under steady

state hemispherical diffusion in PBS buffer. Both AA and UA exhibit well defined oxidation

waves with half-wave potentials around 0.5 V vs Ag/AgCl that correspond to the irreversible

exchange of 2 electrons and 2 protons20,21 at this scan rate. These CVs clearly indicate the

overlapping of the electrochemical oxidation waves of both species at the same potential

range.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 85

Figure 4.2 Cyclic voltammogram of uric acid (dashed line) and ascorbic acid in phosphate

buffer solution scanned at 0.1 V.s-1 indicating overlapping oxidation potential with half-wave

potentials around 0.5 V vs Ag/AgCl.

More precisely, the EC oxidation of UA gives, through reversible 2 electrons-two protons

exchange, an unstable bis-imine that could exist in two tautomeric forms (species IIa and IIb,

figure 4.3).24 This bis-imine compound is then decomposed, at the experimentally used pH of

7.2, into uric acid-4,5-diol (species IV, figure 4.3), through two, fast irreversible hydration

steps.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 86

Figure 4.3 Proposed schematics for UA oxidation in the presence/absence of AA where (I) is

UA, (II) bis-imine compound, (III) imine-alcohol compound and (IV) uric acid-4,5-diol.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 87

Figure 4.4 (a) Cyclic voltammogram of uric acid scanned from 0.1 to 1 V.s-1. When the scan

rate is higher than 0.5 V.s-1, a reduction peak (peak B) appears around -0.1 V vs Ag/AgCl and

(b) cyclic voltammogram of uric acid (dashed line) and ascorbic acid in phosphate buffer

solution scanned at 20 V.s-1.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 88

By increasing the scan rate from 0.1 to 20 V.s-1 , one can see, as already mentioned by

Dryhurst,20 the appearance of a weak reduction peak (for which the amplitude increases with

the scan rate; refer to figure 4.4a) at a potential of -0.1 V vs Ag/AgCl (figure 4.4b). Indeed,

provided the sweep rate is fast enough, the primary oxidized products IIa or IIb can be

electrochemically reduced owing to the reversibility of the first oxidation step due to the

instability of the imine compound. At slower scan rates, oxidation products of UA are rapidly

hydrated to irreversible uric acid-4,5-diol. In contrast to UA and as expected, no reduction

wave of the electrochemically produced dehydroascorbic acid (DHAA) was observed within

the scanned potential window (figure 4.4b), even at the highest investigated scan rate of 20

V.s-1. However, DHAA is known to strongly adsorb at electrochemical interfaces and may

induce diamond fouling along UA determination. Hence, according to the aforementioned

electrochemical behavior, fast CV can be used to selectively determine the concentration of

UA.

Elsewhere, one can deduce from figure 4.2 for UA, the existence of a second flat wave

(oxidation half-wave potential of 0.8V vs Ag/AgCl). Indeed, uric acid-4,5-diol undergoes

subsequent chemical rearrangement at pH 7 leading to the formation of allantoin and urea as

main products. However, parabanic acid can be also produced from uric acid-4,5-diol,

through complex chemical rearrangement to dihydroxyimidazole and subsequent 2 electrons-

2 protons electrochemical oxidation. Indeed, Dryhurst clearly shown that such latter

decomposition follows a minority path which increases in yield with acidity of the media.20,24

Moreover, the yield of this secondary path strongly depends on the electrode material as

reported by Struck et al.25 who detected parabanic acid reduction using polarography

following UA oxidation at spectroscopic graphite electrode in place of pyrolitic graphite. In

such a way, according to the local pH decrease at the vicinity of the electrode owing to UA

oxidation and to the acidic comportment and the nature of hydrogenated diamond electrodes,

parabanic acid may be produced with a yield of around 6 % (obtained from ratio of the

plateau limiting currents that involve both 2 electrons and 2 protons). One can notice, the

disappearance of such a second electrochemical step in fast CV (figure 4.4b), this behavior

can be ascribed to the enhanced electrochemical recycling of UA that decreases the yield of

production of parabanic acid which is produced through a slow chemical and electrochemical

pathway. Moreover, this second electrochemical oxidation is hardly visible due to the Cottrell

behavior of the anodic current of UA primary oxidation wave.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 89

4.3 Calibration curves for UA concentration

Cyclic voltammograms of PBS solution containing uric acid and ascorbic acid are different

when the concentration of ascorbic acid exceeds 1 mM. For lower concentrations of ascorbic

acid (less than 1 mM) an oxidation and a reduction peak were observed and the calibration

process is explained in section 4.3.1. Section 4.3.2 explains the calibration of uric acid at

higher ascorbic acid concentration (more than 1 mM) where two oxidation peaks were

observed. Hence two calibration processes were modeled for low and high concentration of

ascorbic acid and are termed as Model 1 and 2 respectively.

4.3.1 Model 1: Low ascorbic acid concentration

CV of UA (1 mM), at BDD electrode in a solution that contains varying concentrations of

AA, is shown in figure 4.5. In order to plot the calibration curve, the concentration of UA

solution was varied from 0, 500, 1000, 1500, 2000 to 2500 µM and AA concentration from 0,

250 to 500 µM. CVs of different combinations of UA and AA mixtures were carried out and

the peak oxidation current (iA) and peak reduction current (iB) were recorded at 20 V.s-1.

Figure 4.5 Cyclic voltammogram of 1 mM uric acid and ascorbic acid (0, 250 and 500 µM)

in phosphate buffer solution scanned at 20 V.s-1.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 90

Two 3D curves (figure 4.6 and 4.7) were plotted with UA and AA concentrations (CUA and

CAA) on X and Y axis and iA or iB values on Z axis respectively (see Appendix A for more

details). Two second order equations were derived from the curve which corresponds to

Model 1:

iA = 4.24+0.07CUA+0.06CAA-2.9x10-6CUA²-2.36x10-5CAA²-1.82x10-5CUACAA (4.1)

iB = 8.2+0.02CUA-0.009CAA-1.9x10-6CUA²-2.67x10-7CAA²-8.52x10-6CUACAA (4.2)

By solving the equation 4.1 and 4.2, the concentration of UA and AA (CUA and CAA) can be

obtained. It was observed that iA increases with the concentration of UA as well as that of AA

whereas iB decreases with concentrations of AA but increases with concentrations of UA. AA

can deactivate the BDD electrode due to deposition of its oxidation product. It was observed

that the reactivity (k0) of the active BDD electrode was reduced by 10% after few CVs in

solutions containing AA. This could be one explanation for the decrease in iB values when the

concentration of AA is increased. The other assumption is derived from the anti-oxidant

nature of AA. AA is known to reduce quinone imines.26 In the presence of AA, the bis-imines

might have been attacked or quickly reduced to UA chemically. Hence, as the concentration

of AA is increased, the amplitude of iB is decreased.

Figure 4.6 Second order curve fitting results for UA concentration (CUA) at low concentration

of AA (CAA) where iA is the peak oxidation current observed at 20 V.s-1.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 91

Figure 4.7 Second order curve fitting results for UA concentration (CUA) at low concentration

of AA (CAA) where iB is the peak reduction current observed at 20 V.s-1.

4.3.2 Model 2: High ascorbic acid concentration

When the concentration of AA was increased beyond 1mM, the peak B disappeared

completely for UA concentrations below 2 mM. Hence model 1 cannot be used to determine

the concentration of UA. However a third peak (peak C) appears at 0.8 V vs Ag/AgCl as seen

in figure 4.8 when the electrodes where scanned from -0.3 V to 1.4 V. Thus, a second model

(Model 2) is proposed using the two peaks (peak A and C) and their corresponding oxidation

peak currents (iA and iC). For different concentrations of UA and AA mixtures, the peak

oxidation current (iA) and second peak oxidation current (iC) were extracted from the CV

measurements at 20 V.s-1. The concentration of UA solution was varied from 0, 500, 1000,

1500, 2000 to 2500 µM and AA concentration from 1, 3, 5, 7 to 9 mM.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 92

Figure 4.8 Cyclic voltammogram of 1.5 mM uric acid and ascorbic acid (0, 2 and 4 mM) in

phosphate buffer saline solution scanned at 20 V.s-1.

Like Model 1, two 3D curves (figure 4.9 and 4.10) were plotted with UA and AA

concentrations (CUA and CAA) on the X and Y axis and iA or iC values on Z axis respectively

(see Appendix A for more details). The equations of Model 2 are:

iA = 53.48+0.05CUA+0.03CAA-2.56x10-6CUA²-6.69x10-9CAA²-5.75x10-6CUACAA (4.3)

iC = 38.4+0.015CUA+0.02CAA+3.29x10-6CUA²+5.11x10-7CAA²+3.93x10-6CUACAA (4.4)

The concentration of UA and AA (CUA and CAA) can thus be obtained by solving equations

4.3 and 4.4. Both iA and iC increases with the concentration of UA and that of AA. For higher

concentrations of AA (> 5 mM), it was observed that the value iA does not depend much on

UA concentration. As proposed earlier, the AA catalyzes the production of imine-alcohol, no

peak B is observed at higher concentrations of AA. On BDD electrodes, the peak C was not

observed during the CV in solely AA even at higher concentrations.

The simultaneous oxidation of UA in presence of high concentrations of AA presents a

complex mechanism. At higher concentration of AA, the adsorbed oxidation products of AA

influence the electro-kinetics of UA oxidation. A possible explanation is related to the fouling

properties of DHAA. Thereby, due to the blockade of electroactive sites by DHAA, UA

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 93

oxidation may take place through two different pathways: one via still electroactive DHAA-

free diamond-surface sites and one via DHAA blocked sites through the adsorbed species.

Oxidation of UA at the active sites and at the fouled surface could cause the peak separation.

On the other hand, one can note that peak C potential fits quite well with the potential of the

second plateau observed at slow scan rate, thus suggesting the presence of parabanic acid.

Hence, the increase in the peak C amplitude with both UA and AA can be explained first by a

higher generation of parabanic acid with UA concentration and secondly to the possible

recycling of parabanic acid owing to the antioxidant nature of AA. Indeed, some

complementary studies are necessary to examine more in depth the contributions of parabanic

acid and/or DHAA fouling.

Figure 4.9 Second order curve fitting results for UA concentration (CUA) at high

concentration of AA (CAA) where iA is the peak oxidation current observed at 20 V.s-1.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 94

Figure 4.10 Second order curve fitting results for UA concentration (CUA) at high

concentration of AA (CAA) where iC is the second oxidation current observed at 20 V.s-1.

4.4 Proposed model vs Spectrophotometric quantification

Human urine samples were collected from volunteers as real samples for analysis by the

proposed models and were compared to spectrophotometric method. In the

spectrophotometric method, uric acid is oxidized to allantoin in the presence of enzyme

uricase, which leads to formation of H2O2 which reacts with 4-amino phenazone in the

presence of peroxidase to form quinone-diimine. The intensity of the color of the quinone

diimine is directly proportional to the concentration of uric acid. Human urine samples were

neither diluted nor pretreated for EC quantification by the proposed models. CV was done on

each urine sample and, based on the nature of the peaks, using model 1 or model 2; the

equations were solved to identify the UA value.

The results are presented in Table 4.1. The measured values using the proposed model were

observed to be very close to the spectorphotometric results with a maximum difference of

13%. The percentage error was calculated using the following formula:

%E = (P.M – S)/S (4.5)

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 95

where P.M is the values obtained from proposed models and S that of spectrophotometry. The

spectrophotometric measuring technique, although widely used in the field of UA estimation

has an error percentage of 5%. UA and AA are fast antioxidants and the time lag between the

spectrophotometric and EC measurement might have also influenced the difference in the

results. The other parameter which could affect the measurement is the variation in electron

transfer rate k0 of the electrode.

Table 4.1 Comparison of uric acid concentration measured in different urine sample using

model 1 or model 2 and the spectrophotometric technique.

Sample No. Spectrophotometer (mM) Proposed Model (mM) % Error

1 3.66 4.15 13

2 4.76 5.1 7

3 3.01 2.68 -11

4 5.90 5.15 -12

5 2.78 2.74 -1.4

4.5 Automation of quantification procedure and in-situ cleaning

When electrodes are used continuously in a biological fluid, they lose their reactivity because

of fouling27. Electrode fouling can be due to adsorption or adhesion of biomolecules such as

proteins, enzymes, cells, intermediate products of oxidation of organic compounds, etc28.

Although hydrogen-terminated BDD exhibit high reactivity, continuous use in urine leads to

deactivation of electrode reactivity because of fouling. This would lead to difficulty in the

automation of the quantification process. In chapter 2, we demonstrated an in-situ activation

process of BDD electrodes which has been tried on several biological and synthetic fluids.

Human urine samples were diluted 2 fold in PBS solution. From this solution, 5 other

solutions were prepared by adding different quantities of UA (400, 800, 1200, 1600 and

2000µM). CVs at 20 V.s-1 were performed in solutions 1 (diluted urine) to 6 (diluted urine +

2000 µM UA). iA and iB values were obtained from each scan and from the CV of solution 1

the UA concentration was estimated to be 2012 µM. It was observed that the values of iA and

iB were not increased as expected, after each scan (solution 1 to 6) despite the increase in

concentration of UA (figure 4.11). This is due to fouling of the electrode. The electrodes were

EC cleaned and CVs were carried out from solution 1 (diluted urine) to 6 (diluted urine +

2000 µM UA). Between each CV the electrodes were activated in the previous solution. The

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 96

EC activation was performed directly in the solution containing urine, by applying a train of

negative current pulses consisting of 3 pulses where each pulse had an amplitude of -100

mA.cm-2 and duration of 100 ms and duty cycle of 90 % (refer to chapter 2 section 2.6). The

values of iA and iB increased after each scan as expected (figure 4.12).

Figure 4.11 Cyclic voltammogram of urine diluted by 2 fold and that of diluted urine

containing added uric acid (250 – 1250µM) scanned at 20 V.s-1.

Values of iA and iB were extracted from these graphs and are denoted iAO1, iBO1 and iAO2, iBO2

for trails with and without activation in between two CVs. Also the theoretical values of iA

and iB were calculated using the models and were denoted iAC and iBC. Comparison between

the calculated value iAC and the observed values iAO1 (trials with activation) and iAO2 (trials

without activation) shows that there is a negligible difference between iAC and iAO1 when

compared to the difference between iAC and iAO2. The percentage difference between

calculated and observed values was estimated using the following formula:

E = (iC-iO) /iC (4.6)

with EA1 the percentage difference between iAC and iAO1 and EA2 the percentage difference

between iAC and iAO2, EB1 is the percentage difference between iBC and iBO1 and EB2 is the

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 97

percentage difference between iBC and iBO2 . A detailed comparison between the different

values is depicted in table 4.2 and figure 4.13.

Figure 4.12 Cyclic voltammogram of urine diluted 2 fold and that of diluted urine containing

added uric acid (250 – 1250µM) scanned at 20 V.s-1. The electrodes were activated

electrochemically in the same solution in between two successive scans.

Table 4.2 Comparison between the calculated values (iAC, iBC), observed values with

activation between CVs (iAO1, iBO1) and observed values without activation between CVs

(iAO2, iBO2) for the peak currents iA and iB extracted from CV in urine solutions 1 to 6 where

EA1, EA2, EB1 and EB2 are percentage differences of observed values from calculated values.

UA Conc.

(µM)

iAC

(nA)

iAO1

(nA)

EA1

(%)

iAO2

(nA)

EA2

(%)

iBC

(nA)

iBO1

(nA)

EB1

(%)

iBO2

(nA)

EB2

(%)

2012 138 138 0 138 0 -25 -25 0 -25 0

2412 157.3 162 -2.99 150 4.64 -27.06 -26 3.92 -26 3.92

2812 176.83 182 -2.92 156 11.77 -28.54 -27 5.39 -25 12.40

3212 194.85 200 -2.64 161 17.37 -29.39 -28 4.72 -25 14.94

3612 211.95 210 0.92 174 17.90 -29.63 -28.5 3.81 -25 15.63

4012 228.1 228 0.04 172 24.59 -29.95 -30 -0.17 -24 19.87

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 98

Figure 4.13 Comparison between the calculated value (iAC, iBC), observed value with

activation between CVs (iAO1, iBO1) and observed value without activation between CVs (iAO2,

iBO2) for the peak currents iA and iB extracted from CV in urine solutions 1 to 6.

Due to heavy fouling of the electrode, the electrode was deactivated and hence the value of

EA2 is as high as 25%. As the electrode is not activated after each CV, the value of EA2

increased after each scan. In contrast, the value of EA1 is less than 3% indicating the closeness

of calculated values and observed values. The same is the case for EB1 and EB2 with a

maximum EB1 of 5.4% and EB2 of 20%. It is clearly demonstrated that a simple EC activation

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 99

procedure can restore the lost reactivity of the electrode and the electrode can be reused in the

sample solution for further measurements.

Conclusions

Selective determination of UA in the presence of AA was achieved using a BDD

microelectrode without any further modification. Comparison of the EC quantification

technique and the spectrophotometric technique shows that an accurate measurement can be

carried out using the 2 proposed models. This technique highlights the potential of BDD

microelectrodes as electro-analytical sensors owing to their low double layer capacitance,

robustness at high current density and corrosion resistance. The EC treatment retrieves the

lost reactivity of an electrode fouled by urine without using any specific reagent or solution.

The advantage of this technique is that it can enhance the reusability of the BDD

microelectrode by activating in urine itself. This demonstrates the possibility of automation of

UA quantification as the electrode can be activated directly in urine and hence it can be used

for continuous monitoring for long period of time. The time taken for activation is 300 ms and

the time taken for CV at 20 V.s-1 is less than 200 ms.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 100

Bibliography

(1) Álvarez-Lario, B.; Macarrón-Vicente, J. Rheumatology (Oxford, England) 2010, 49, 2010–2015.

(2) Popa, E.; Kubota, Y.; Tryk, D. A.; Fujishima, A. Analytical chemistry 2000, 72, 1724–1727.

(3) Amir, W.; Mohammad, Y.; Abdul, N.; Ghulam, M.; Izhar, H. Chemical Research in Chinese Universities 2011, 27, 929–933.

(4) Martinon, F.; Pétrilli, V.; Mayor, A.; Tardivel, A.; Tschopp, J. Nature 2006, 440, 237–241.

(5) Daoussis, D.; Kitas, G. D. Rheumatology (Oxford, England) 2011, 50, 1354–1355.

(6) Sun, Q.; Shen, Y.; Sun, N.; Zhang, G. J.; Chen, Z.; Fan, J. F.; Jia, L. Q.; Xiao, H. Z.; Li, X. R.; Puschner, B. European journal of pediatrics 2010, 169, 483–489.

(7) Shaikh, A.; Wiisanen, M. E.; Gunderson, H. D.; Leung, N. The Annals of pharmacotherapy 2009, 43, 1370–1373.

(8) Dinour, D.; Gray, N. K.; Campbell, S.; Shu, X.; Sawyer, L.; Richardson, W.; Rechavi, G.; Amariglio, N.; Ganon, L.; Sela, B.-A.; Bahat, H.; Goldman, M.; Weissgarten, J.; Millar, M. R.; Wright, A. F.; Holtzman, E. J. Journal of the American Society of Nephrology竺: JASN 2010, 21, 64–72.

(9) Lameire, N.; Van Biesen, W.; Vanholder, R. Nephrology, dialysis, transplantation竺: official publication of the European Dialysis and Transplant Association - European Renal Association 1999, 14, 2570–2573.

(10) Li, Y.; Lin, X. Sensors and Actuators B: Chemical 2006, 115, 134–139.

(11) Lin, X.; Zhang, Y.; Chen, W.; Wu, P. Sensors and Actuators B: Chemical 2007, 122, 309–314.

(12) Lin, L.; Chen, J.; Yao, H.; Chen, Y.; Zheng, Y.; Lin, X. Bioelectrochemistry (Amsterdam, Netherlands) 2008, 73, 11–17.

(13) Zen, J.; Tang, J. Anal. Chem. 1995, 67, 1892–1895.

(14) Zen, J.-M.; Chen, P.-J. Analytical Chemistry 1997, 69, 5087–5093.

(15) Wang, Z.; Wang, Y.; Luo, G. The Analyst 2002, 127, 1353–1358.

(16) Fernandez, L.; Carrero, H. Electrochimica Acta 2005, 50, 1233–1240.

(17) Nakaminami, T.; Ito, S.; Kuwabata, S.; Yoneyama, H. Analytical chemistry 1999, 71, 1928–1934.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 101

(18) Strochkova, E. M.; Tur’yan, Y. I.; Kuselman, I.; Shenhar, a Talanta 1997, 44, 1923–8.

(19) Safavi, A.; Maleki, N.; Moradlou, O.; Tajabadi, F. Analytical biochemistry 2006, 359, 224–9.

(20) Dryhurst, G. Journal of electrochemical society 1972, 119, 1659–1664.

(21) Hu, I.-F.; Kuwana, T. Analytical chemistry 1986, 58, 3235–3239.

(22) Lawrence, N. S.; Pagels, M.; Meredith, A.; Jones, T. G. J.; Hall, C. E.; Pickles, C. S. J.; Godfried, H. P.; Banks, C. E.; Compton, R. G.; Jiang, L. Talanta 2006, 69, 829–834.

(23) Konopka, S. J.; Mcduffie, B. Analytical chemistry 1970, 42, 1741–1746.

(24) Fry, A. J.; Dryhurst, G. Organic Electrochemistry; Springer-Verlag, 1972; p. 85.

(25) Struck, W. A.; Elving, P. J. Analytical chemistry 1964, 36, 1374–1375.

(26) Kang, P.; Dalvie, D.; Smith, E.; Renner, M. Chemical research in toxicology 2009, 22, 106–117.

(27) Iniesta, J.; Michaud, P. A.; Panizza, M.; Comninellis, C. Electrochemistry Communications 2001, 3, 346–351.

(28) Wang, J.; Hutchins, L. D. Analytical chemistry 1985, 57, 1536–1541.

Diamond microelectrodes: Electroanalytical application

Chapter IV Page 102

CHAPTER V

Diamond microelectrodes: Electrophysiological

applications

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 104

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 105

Introduction

Electrophysiology is the study of electrical properties of cells and tissues which involve

measurement of the voltage change of a biological entity. It is a powerful approach not only to

study the electrical activity of animal cells to understand the working of the nervous system,

brain, hypothalamus, etc. but also to diagnose and treat nervous system disorders. The

electrical activity of neurons can be measured directly using extracellular microelectrode

arrays. Using the microelectrodes the potential changes in the vicinity of the electrode, caused

by currents flowing across neuronal membranes of multiple neurons, can be detected. Neural

recording can be performed both in-vivo1–3 (on live animal) and in-vitro4–6 (extracted cells).

Electrical stimulation of nerve cells is widely employed in neural prostheses, clinical therapies

and neuroscience studies, as it has the potential to excite virtually every tissue. It is highly

significant for individuals to restore the senses of hearing7 and vision8 and in the treatment of

Parkinson disease,9 Paraplegia,10 etc. An implanted multichannel microelectrode array can be

used to transmit electrical signals to the neurons and thus modulate their behavior. The

electrodes should be biocompatible, micro-structured, deliver high charge injection limit

without degradation and corrosion resistant.

Boron-doped diamond films, with their morphological and microstructural stability at high

anodic current densities11 and wide electrochemical potential window in aqueous

electrolytes12, are ideal candidates for neural recording and stimulation. Several studies have

been conducted on diamond microelectrodes for neuro-chemical and neuro-electrical

recording.13–15 The biocompatibility of diamond surfaces has been investigated by our group

by implanting BDD microelectrodes in rat’s eyes and no major gliosis was detected, which is

commonly used as a biocompatibility indicator.16 Another advantage of diamond is the weak

adsorption of polar molecules on its surface.17 As neural recording is done in ion and protein-

rich environments which cause electrode fouling and hence to a decrease in the signal-to-

noise ratio (SNR), a foul-resistant and chemically stable electrode is required.

The MEAs fabricated using the technique described in chapter 3 (design 2) were used for in-

vitro electrophysiological measurements. Additionally we describe the fabrication and

characterization process of a neural prosthesis: a retinal implants using BDD MEAs. Studies

on the improvement of the charge injection limit are also included. This was achieved by

nano-structuring the electrode.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 106

5.1 Electrophysiological characterization of MEA

The 8x8 UMEA (diameter of the electrode = 14 µm) was characterized (refer to chapter 3

section 3.1.2) in collaboration with Prof. Blaise Yvert from INCIA, Bordeaux.

Electrophysiological characterization included impedance measurement and noise level

calculation.

5.1.1 Impedance measurement

Impedance measurements at 1 kHz (sine wave) were performed using either an IMP-I

Electrode Impedance Tester from Bak Electronics Inc (Mount Airy, USA) or a NanoZ device

from Multichannel Systems (Reutlingen, Germany, refer figure 5.1 ) in 0.1 0.5 M KCl

aqueous solution.18 Figure 5.2 a and b shows the magnitude and phase of the impedance value

obtained. The red spots in those figures shows abnormal value (typically very low magnitude

and phase angle) and are omitted from the estimation of average impedance of the array. Also

the green spots correspond to the electrodes that exhibited reduced electrochemical window

(refer to chapter 3 section 3.3.1) and were also omitted from the calculation (associated with

non-continuous diamond film).

Figure 5.1 8x8 BDD UMEA fixed on NanoZ device to measure the magnitude and phase of

the impedance at 1 kHz.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 107

(a)

(b)

Figure 5.2 (a) Magnitude and (b) phase of the impedances at 1 kHz for all 59 boron doped

diamond ultra-microelectrodes of 14 µm diameter. The red spots correspond to those

electrodes with abnormal magnitude and phase and the green spots correspond to those

electrodes which did not exhibit typical diamond potential window.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 108

The impedances (Z) of other electrodes of this array measured at 1 kHz, were found to be

very homogenous around 1033 Ʋ 50 kohm, (mean Ʋ sd). One can observe from figure 5.2 b

that the impedance measured is purely reactive as the phase angle is almost -90°. The double

layer capacitance obtained from the Nyquist plot (chapter 2 section 3.3.3) for the electrodes

were 150 pF and at 1 kHz the impedance is 1061 kohm (equation 2.1) which shows a very

good correlation to the observed average value (of 1033 kohm).

傑 噺 怠態訂捗寵呑 (5.1)

where f is the frequency of the sinusoidal wave and CD is the double layer capacitance.

5.1.2 Noise level measurement

To measure the intrinsic noise level of the electrodes, the electrical potential was recorded for

1 min in physiological liquid between each of the 60 microelectrodes and an Ag/AgCl ground

electrode pellet. Signals were 1100 x amplified and band-pass filtered between 1 Hz and 3

kHz using MCS MEA1060-Up-BC filter amplifiers from Multichannel Systems (Reutlingen,

Germany). Data were acquired at 10 kHz using two synchronized CED Power1401 AD

converters and the Spike2 v6 software from Cambridge Electronic Design (Cambridge,

England). The standard deviation of the signal s was then calculated over the 1 min

recording period for each electrode of the array. Because this noise level was composed of

both the intrinsic noise level of the electrodes e and the electronic noise level of the

amplifiers a, we assumed statistical independence of these two noise sources and estimated

the intrinsic noise level of each electrode as:

22ase (5.2)

where a = 1.4 µV was measured with the amplifier inputs connected to the ground. The

average noise level of the electrodes was observed to be 6 µV. This value is significantly

higher when compared to commonly used implants like titanium nitride.19

5.1.3 Neural recording

Embryos were surgically removed from pregnant mice and the hindbrain-spinal cords were

dissected in cooled Ringer solution (pH 7.5) composed of (in mM): 113 NaCl, 4.5 KCl, 2

CaCl2.2H2O, 1 MgCl2.6H2O, 25 NaHCO3, 1 NaH2PO4.H2O and 11D-Glucose gassed with

carbogen (95% O2 and 5% CO2).20 The dissected hindbrain-spinal cords were placed dorsally

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 109

in the MEA cylindrical chamber, with the external side of the neural tube in contact with the

microelectrodes of the array (refer figure 5.3). The neural tissue was stabilized using a plastic

net with small holes (70 µm x 70 µm) in order to achieve a tight and uniform contact with the

microelectrodes. Experiments were conducted at room temperature. For recording and

stimulating the neurons, a 64 channel electronic device called BioMEAƚ (from Biologic) has

been used and the whole experimental setup was mounted on it.21

Figure 5.3 A whole mouse embryonic hindbrain-spinal cord system positioned on an 8 x 8

microelectrode array.

Figure 5.4 A close-up view of wave propagation triggered by electrical stimulation on 1

channel and waves of spontaneous activity on 24 channels

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 110

The spontaneous activity of the developing neural network was recorded with the help of the

microelectrodes. Each spike lasts for a few seconds and recurs every few minutes. The

spontaneous spikes generated by the neural network change with the development in the

embryonic stage. For the establishment of neural connectivity it is essential to have this

spontaneous neural activity.20 This activity is due to intracellular calcium influxes which play

an important role in promoting axonal growth. The spontaneous spikes generated exhibit

typical amplitude of 20 µV.

The BioMEAƚ system was used to stimulate neurons electrically through one microelectrode

with biphasic current stimuli (Ʋ100 µA, 500 µs/phase) with an external reference electrode

(Ag/AgCl). All other electrodes are used to record the stimulated activity of the neuron. Each

stimulus triggered an episode of activity which propagated rostrocaudally along the whole

preparation. Figure 5.4 shows the stimulated and spontaneous activity of the neurons.

There are several possible reasons for the low SNR.15 If the probe position is not close enough

to the neurons, then the strength of the neuron signal attenuates with increasing distance

between the recording site and the target neurons. When compared to metal electrodes, the

resistivity of boron doped-diamond is relatively high and this causes an increased thermal

noise. A novel fabrication technique is developed and is described in section 5.3 to decrease

the impedance and increase the SNR.

5.2 Diamond microelectrode array as neural prosthesis: Retinal implants

The neural interface and prosthesis presents many significant challenges in the development

of advanced devices designed to restore function in neurologically impaired patients.

Implantable microelectrodes can be used to record neuronal action potentials or local field

potentials and stimulate the neurons. European (NEUROCARE, DREAMS) and French

(MEDINAS, IMPLANT) research projects aim to develop better retinal implants based on

carbon materials. Diamond, being robust and bio-inert, has been chosen as a candidate for

recording and stimulating the neurons. Age-related macular degeneration, retinis pigmentosa,

etc. can lead to partial or complete blindness and despite enormous advancement in clinical

treatments, there are no methods to prevent or cure these diseases.22 The development in

micro-fabrication techniques have opened new possiblity for developing microelectrode

arrays that can be implanted in the retina and can stimulate electrically the neural network to

restore lost function.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 111

5.2.1 Platinum microelectrode arrays

Implantable platinum microelectrode arrays were fabricated as well as the boron-doped

diamond MEA in order to compare their performances in the electrochemical,

electrophysiological and biocompatibility tests. Figure 5.5 describes the technological

processes used in the fabrication of Pt MEAs. Four inch silicon wafers were oxidized to form

a 500 nm silicon dioxide layer (refer to chapter 3 section 3.1 for more details). The oxide

layer acts as the sacrificial layer. A 10 µm layer of polyimide precursor was applied on top of

the oxide layer by spin coating and was cured at 375 °C under nitrogen. The polyimide was

etched selectively under oxygen plasma using a metal etch mask. Pt was deposited by

physical vapor deposition over the patterned polyimide layer and was patterned using a lift-off

process. A second polyimide layer of 10 µm thickness was deposited and the contact pads

were exposed by selective etching of the polyimide. The polyimide was then detached from

the silicon substrate by dissolution in HF.

Figure 5.5 Schematics of implantable 8 x 8 Pt microelectrode array fabrication process.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 112

The polyimide coated Pt retinal implants have an implant “head” dimension of 1.8 mm in

diameter and rectangular “tail” of 27 mm long and 1.6 mm wide. Figure 5.6 shows the photos

of 8x8 electrode array. The electrodes are of 14 µm in diameter and inter-electrode distance of

100 µm.

Figure 5.6 Pt soft implant (8 x 8 electrode array)

The Pt microelectrode arrays were characterized to evaluate their electrochemical

performances. Detailed characterization techniques are described in chapter 1 (section 1.4)

and 3 (section 3.3). Typically the electrochemical window recorded for the Pt electrodes was

1.8 V (measured in in 0.5 M aqueous LiClO4 solution) and hence close to the values reported

by other groups.23 Figure 5.7 shows a voltammogram of one of the microelectrode (14 µm in

diameter) scanned at 100 mV.s-1.

Figure 5.7 Cyclic voltammetry (scan rate=100 mV.s-1) of Pt microelectrode in 0.5 M

aqueous LiClO4 solution

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 113

Figure 5.8 Nyquist plot (experimental data and fitted data) of Pt microelectrode.

Figure 5.9 Cylcic voltammogram of Pt microelectrode at 100 mV.s-1 in 0.5 M

potassium chloride aqueous solution containing 1 mM Fe(CN)64- ion.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 114

The steady state limiting current (in 1mM Fe(CN)63-/4- in 0.5 M KCl solution at 100 mV.s-1 -

figure 5.9) was observed to be 1.5 nA (compared to the theoretical value of 1.8 nA). The

electron transfer rate k0 and the double layer capacitance CD estimated from Nyquist plot

(figure 5.8) were 0.003 cm.s-1 and 126 pF respectively. Although the microelectrodes exhibit

poor electron transfer rates and limiting current, they find application in electrophysiological

applications due to its high double layer capacitance that reduces the noise level and increases

the charge injection limit.15,24

5.2.2 Diamond microelectrode arrays

Implantable BDD microelectrode arrays were fabricated using a similar process to that

described in section 5.2.1 and figure 5.10. Diamond nanoparticles were fixed on an oxidized

silicon wafer and were patterned using a metal mask (refer to chapter 3 section 3.1). After the

diamond growth, a polyimide precursor was applied on top of the oxide layer by spin coating

and was cured. The polyimide was etched selectively and Ti/Pt contacts and tracks were

deposited by physical vapor deposition over the patterned polyimide layer and are in contact

with BDD film. A second polyimide layer of 10 µm thickness was deposited and the contact

pads were exposed by selective etching of the polyimide which is followed by substrate

detachment.

Figure 5.10 Schematics of implantable 8 x 8 BDD microelectrode array fabrication process.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 115

The electrochemical window recorded for the BDD electrodes was typically 3 V and is less

than the values reported by other groups.25 Figure 5.11 shows a voltammogram for a BDD

microelectrode (14 µm in diameter) scanned at 100 mV.s-1. The peak current (in 1mM

Fe(CN)63-/4- in 0.5 M KCl solution at 100 mV.s-1.- figure 5.12) was observed to be 320 nA

which is 178 times higher than the theoretical value (assuming a steady state voltammogram).

From the Nyquist plot (figure 5.13), k0 and CD were estimated and were 2.1 cm.s-1 (RT = 84

kohm) and 24 nF (again the values are much higher than the expected values). The high

electron transfer rate, capacitance and limiting current suggest that there might be some

leakage during the fabrication process where the electrodes are shorted or in contact with the

grid (which is used to achieve focal stimulation). Also the reduced window suggests that

titanium or the electrolyte itself might have diffused through the polyimide and thus

demonstrates the difficulty of structuring a coherent process here.

Figure 5.11 Cyclic voltammetry (scan rate=100 mV.s-1) of BDD microelectrode in 0.5

M aqueous LiClO4 solution

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 116

Figure 5.12 Nyquist plot (experimental data and fitted data) of BDD microelectrode.

Figure 5.13 Cylcic voltammogram of BDD microelectrode at 100 mV.s-1 in 0.5 M

potassium chloride aqueous solution containing 1 mM Fe(CN)64- ion.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 117

The technological process has to be improved in order to fabricate implantable BDD

microelectrode array with no leakage. Weiland et al. have evaluated the polymer layers:

polyimide, parylene and silicone as the retinal prosthesis substrates.26 The study indicates that

parylene possess the best property among the polymers and multi-polymer approach is best

suited for the implants. A long term soak-test is necessary to evaluate the insulation property

of the polymer layer.

5.3 Nanograss diamond MEA

For neural stimulation, electrodes should be capable of injecting relatively large currents

while minimizing electrode degradation due to faradaic effects. In order to allow long-term

operation, the surface of the microelectrodes should be able to facilitate charge transport

without degradation of the electrode. This led us to the fabrication of BDD nanograss

microelectrode arrays with enhanced double layer capacitance ideal for neural recording and

stimulation. Since the surface was not modified, rather they were nanostructured, the

electrodes possessed the physical and chemical properties of diamond with superior charge

injection limits.

5.3.1 Fabrication of nanograss MEA

There has been several techniques reported for nanostructuring diamonds which including

oxygen plasma etching through porous anodic alumina fims,27 self-aligned Au nanodots as an

etching mask in hydrogen/argon plasma,28 nanodiamond powder hard mask in reactive ion

etching (RIE) in O2/CF4 gas mixture29 etc. We have adopted a simple method to nanostructure

BDD MEAs (4 x 16 electrodes of diameter varying from 10 to 70 µm, refer chapter 3 section

3.1.2 for detailed fabrication process) by using RIE without using any mask.30 Diamond was

etched by the use of oxygen plasma under a pressure of 8 mbar at 30 sccm and plasma power

of 200 W for 1 minute. As per Wei et al. the boron dopant atoms in the diamond act as the

mask during plasma etching and the boron oxides are redeposited on top of the nanograss and

continues to serve as the mask.30 Even if this mechanism is not well understood, we used the

technique to successfully fabricate BDD nanograss.

5.3.2 SEM characterization

Figure 5.14 shows the SEM images of a nanostructed BDD nanograss electrode (electrode

diameter = 20 µm). The nanograss structures exhibit dimensions of around 100 nm in length

and 10-20 nm in diameter and 20-30 nm of inter-nanograss distance.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 118

Figure 5.14 SEM image of cross-section of BDD nanograss electrode.

5.3.3 Electrochemical characterization

The Nyquist plot obtained was a semicircular spectrum which matches a Randles circuit with

no Warburg impedance component. The values of each component of the system (RS, CD, and

RT) were calculated from these spectra (refer chapter 1 section 1.4.2 for detailed experimental

setup). Figure 5.15 shows the measured and calculated Nyquist plot of one such electrode (of

diameter 20 µm). The resistance measured between the working electrode and the reference

electrode, the ohmic resistance RS, includes the resistance of the solution and the resistance of

the metallic track and was 2076 っ. The fitted CD, and RT values were of 288 pF and 73.73

Mohm respectively.

The double layer capacitance of the nanograss microelectrode was about 30 times higher than

the theoretical value of diamond electrode (3 µF.cm-2) indicating that the effective surface

area of the electrode has been increased by a factor more than 30. The electrode electron

transfer rate k0 was calculated to be 1.15 x 10-3 cm.s-1. Although H-terminated BDD

electrodes are known to exhibit reactivity as high as 0.2 cm.s-1, the oxygen plasma treatment

has reduced the reactivity considerably. Low k0 value is an added advantage to the nano-

structured MEAs as they reduce the faradaic current which involves in electrochemical

changes to the solution while stimulating.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 119

Figure 5.15 Nyquist plot (experimented and fitted curves) of a nanograss diamond

microelectrode, showing the Re (Z) vs - img (Z) and is a semi-circular spectrum indicating

that the straight line (Warburg element) is absent.

Figure 5.16 shows the EC potential window of the nanograss microelectrode which is about

3.4 V (similar to typical BDD window) with reduction and oxidation of water occurring at -

1.5 V and 1.9 V respectively. The CV curves indicate that there is neither crack nor pin-holes

in the passivation layer and that the only current component within the window is the transient

current which charges and discharges the double layer constituting the background current

within the potential window.

The enhanced background current together with the wide potential window increases the

charge injection limit needed for neural stimulation. Steady state limiting current ilim of

ferrous oxidation was calculated from the CV (figure 5.17) in Fe(CN)63-/4- solution and is 8.78

nA. The theoretical value ilim for this 20 µm diameter (with the effective area enhanced by

30.5 times) is 78.53 nA.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 120

Figure 5.16 Cyclic voltammogram of nanograss diamond microelectrode scanned at 0.2 V.s-1

in 0.5 M LiClO4 aqueous solution.

Figure 5.17 Cyclic voltammogram of nanograss diamond microelectrode scanned at 0.1 V s-1

in 0.5 M KCl aqueous solution containing 1 mM of Fe(CN)63-/4-.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 121

The huge difference in the observed and theoretical value of the limiting current can be

explained by the low reactivity of the nanograss microelectrodes, probably due to the fact that

not the entire surface is active due to high series resistance of the diamond nanowires.

5.3.4 Electrophysiological characterization

Impedance measurements at 1 kHz on a 4 x 16 array of BDD nanograss of varying diameter is

depicted in figure 5.18. The impedances of the electrodes of this array measured at 1 kHz,

were not very homogenous around 202 Ʋ 139 Mohm.µm2 (mean Ʋ sd). The impedance

measured is reactive and the phase angle is -86°. The double layer capacitance obtained from

the Nyquist plot for the electrode (20 µm diameter) was 288 pF and at 1 kHz the impedance is

174 Mohm.µm2. Since the surface of the nanograss is rough, the effective surface area is not

known and hence this can cause variations in the impedance values measured from electrode

to electrode.

Figure 5.18 Electrode impedance measured for the 4 x 16 array.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 122

The noise measured for this array is in good correlation with the impedance. As the diameter

increases the noise level decreases. Figure 5.19 shows the noise level measured for the 4 x 16

MEA. The charge injection limit measured was above 100 µC.cm-2.

Figure 5.19 Electrode impedance measured for the 4 x 16 array.

The BDD nanograss microelectrode array with its very low impedance and noise level and

high charge injection limit could be a very useful tool for electrophysiological applications.

The surface being diamond, the biocompatibility and corrosion resistivity of the electrodes are

expected to be excellent.

Conclusions

Diamond is an excellent neural prosthesis with its superior physical and chemical properties

along with biocompatibility. In this chapter we have seen reported several approaches

proposed to enhance the electrical properties of diamond to make it an ideal electrode for

electrophysiological applications. Although BDD-PPy electrodes cannot be successfully used

for neural prosthesis it has opened ways to other research areas such as the power storage and

transmission.

BDD nanograss microelectrodes exhibit very high charge injection limits and when compared

to titanium nitride and iridium oxide, diamond possess better biocompatibility and

microstructural stability. Increasing the number of electrodes, designing the appropriate

electrode shapes (so far the shapes were all planar) and further electrode surface modification

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 123

can make BDD nanograss MEAs superior devices than other competing techniques for neural

prosthesis.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 124

Bibliography

(1) Blau, a; Ziegler, C.; Heyer, M.; Endres, F.; Schwitzgebel, G.; Matthies, T.; Stieglitz, T.; Meyer, J. U.; Göpel, W. Biosensors & bioelectronics 1997, 12, 883–92.

(2) Cheung, K. C.; Renaud, P.; Tanila, H.; Djupsund, K. Biosensors & bioelectronics 2007, 22, 1783–90.

(3) Kuperstein, M.; Whittington, D. a IEEE transactions on bio-medical engineering 1981, 28, 288–93.

(4) Gross, G. W.; Rieske, E.; Kreutzberg, G. W.; Meyer, A. Neuroscience Letters 1977, 6, 101–105.

(5) Berdondini, L.; van der Wal, P. D.; Guenat, O.; de Rooij, N. F.; Koudelka-Hep, M.; Seitz, P.; Kaufmann, R.; Metzler, P.; Blanc, N.; Rohr, S. Biosensors & bioelectronics 2005, 21, 167–74.

(6) Gross, G. W. 1979, 273–279.

(7) Grill, W. M.; Mortimer, J. T. IEEE Engineering in Medicine and Biology Magazine 1995, 14, 375–385.

(8) Normann, R. a; Maynard, E. M.; Rousche, P. J.; Warren, D. J. Vision research 1999, 39, 2577–87.

(9) Limousin, P.; Krack, P.; Pollak, P.; Benazzouz, A.; Ardouin, C.; Hoffmann, D.; Benabid, A. The New England Journal of Medicine 1998, 339, 1105–1111.

(10) Brindley, G. S.; Polkey, C. E.; Rushton, D. N. Paraplegia 1979, 16, 428–37.

(11) Chen, Q.; Granger, M. C.; Lister, T. E.; Swain, G. M. J. Electrochem 1997, 144, 3806–3812.

(12) Panizza, M.; Michaud, P. a.; Cerisola, G.; Comninellis, C. Journal of Electroanalytical Chemistry 2001, 507, 206–214.

(13) Park, J.; Quaiserova-mocko, V.; Peckova, K.; Galligan, J. J.; Fink, G. D.; Swain, G. M. Diamond and Related Materials 2006, 15, 761–772.

(14) Halpern, J. M.; Xie, S.; Sutton, G. P.; Higashikubo, B. T.; Chestek, C. a.; Lu, H.; Chiel, H. J.; Martin, H. B. Diamond and Related Materials 2006, 15, 183–187.

(15) Chan, H.; Aslam, D. M.; Member, S.; Wiler, J. A.; Casey, B. journal of microelectromechanical systems 2009, 18, 511–521.

(16) Bongrain, A.; Bendali, A.; Lissorgues, G.; Rousseau, L.; Yvert, B.; Scorsone, E.; Bergonzo, P.; Picaus, S. DTIP proceeding 2011, 11–14.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 125

(17) Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, S.; Show, Y.; Swain, G. M. Diamond and Related Materials 2003, 12, 1940–1949.

(18) Prasad, A.; Sanchez, J. C. Journal of neural engineering 2012, 9, 026028.

(19) Ryynänen, T.; Kujala, V.; Ylä-Outinen, L.; Korhonen, I.; Tanskanen, J. M. a.; Kauppinen, P.; Aalto-Setälä, K.; Hyttinen, J.; Kerkelä, E.; Narkilahti, S.; Lekkala, J. Micromachines 2011, 2, 394–409.

(20) Yvert, B.; Branchereau, P.; Meyrand, P. Journal of neurophysiology 2004, 91, 2101–9.

(21) Charvet, G.; Rousseau, L.; Billoint, O.; Gharbi, S.; Rostaing, J.-P.; Joucla, S.; Trevisiol, M.; Bourgerette, A.; Chauvet, P.; Moulin, C.; Goy, F.; Mercier, B.; Colin, M.; Spirkovitch, S.; Fanet, H.; Meyrand, P.; Guillemaud, R.; Yvert, B. Biosensors & bioelectronics 2010, 25, 1889–96.

(22) Meyer, È. Sensors and Actuators A 2002, 97-98, 1–9.

(23) Braga, N. a.; Cairo, C. a. a.; Matsushima, J. T.; Baldan, M. R.; Ferreira, N. G. Journal of Solid State Electrochemistry 2009, 14, 313–321.

(24) Cheung, K. C. Biomedical microdevices 2007, 9, 923–38.

(25) Colley, A. L.; Williams, C. G.; D’Haenens Johansson, U.; Newton, M. E.; Unwin, P. R.; Wilson, N. R.; Macpherson, J. V. Analytical chemistry 2006, 78, 2539–2548.

(26) Weiland, J. D.; Humayun, M. S.; Eckhardt, H.; Ufer, S.; Laude, L.; Basinger, B.; Tai, Y.-C. Conference proceedings竺: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 2009, 4140–3.

(27) Masuda, H.; Watanabe, M.; Yasui, K.; Tryk, D.; Rao, T.; Fujishima, A. Advanced Materials 2000, 12, 444–447.

(28) Zou, Y. S.; Yang, Y.; Zhang, W. J.; Chong, Y. M.; He, B.; Bello, I.; Lee, S. T. Applied Physics Letters 2008, 92, 053105.

(29) Yang, N.; Uetsuka, H.; Osawa, E.; Nebel, C. E. Angewandte Chemie (International ed. in English) 2008, 47, 5183–5.

(30) Wei, M.; Terashima, C.; Lv, M.; Fujishima, A.; Gu, Z.-Z. Chemical communications (Cambridge, England) 2009, 3624–6.

Diamond microelectrodes: Electrophysiological applications

Chapter V Page 126

CHAPTER VI

Conclusions and Future Perspectives

Conclusions and future perspectives

Chapter VI Page 128

Conclusions and future perspectives

Chapter VI Page 129

The global aim of the research discussed in this thesis is the design, fabrication,

characterization and applications of Diamond microelectrodes. Yet the thesis also explores

other topics such as in-situ activation technique, supercapacitors, etc.

4.1 Conclusions

The electrochemical activation process retrieves the lost reactivity of an electrode either aged

in air or fouled by a medium. The other advantage of this technique is to enhance the

reusability of the BDD electrode. Contrary to the other pre-treatments techniques commonly

reported in the literature, for instance anodic, cathodic and thermal, this novel electrochemical

treatment is relatively simpler, fast, and minimum resource consuming. To ensure reliable and

reproducible results, especially when the electrode has not been used for a long period of

time, it is recommended to perform this electrochemical activation prior to the experiments.

Application of the activation technique has been seen in chapter 3 and 4 for cleaning the

electrode surface from residual photoresist and to develop an automated human uric acid

(UA) quantification sensor with in-situ cleaning.

A novel microfabrication technique to fabricate electrochemical and electrophysiological

sensors has been developed for electro-analysis and neural recording. After experimenting

several fabrication designs, a high-yield, reproducible design (design 2) was adopted, where

diamond is grown before the deposition of metal tracks and the passivation layer. The

electrodes exhibited very low leakage current and very high reactivity ideal for

electrochemical sensors.

The strip microelectrodes fabricated as per design 2 were used to develop biosensors. Uric

acid sensor based on the two proposed models indicates that the technique developed can be

used as an alternative quantification process for spectrophotometric measurements. The fast

scan cyclic voltammetry not only separates the uric acid and ascorbic acid peaks, but also

decrease the effective quantification time. As the detection technique do not rely on any

enzymatic reaction and does not need any surface modifications, a single electrode can be

used for several times followed by the electrochemical activation. The in-situ activation

protocol indicates the possibility to incorporate the biosensor to a bed-side monitor as the

process can be fully automated.

Diamond, being biocompatible and possessing very high corrosion resistance, is a very good

candidate for neural prosthesis as well as for in-vivo/in-vitro neural measurements. The

Conclusions and future perspectives

Chapter VI Page 130

microelectrode arrays, although possess features better than conventional metal electrodes, are

inferior compared to titanium nitride and iridium oxide in terms of charge injection limit and

impedance. Attempts have made to address these disadvantages which include BDD surface

modification and nano-structuring. BDD surface modification by BDD-PPy composite had

serendipitously led to the invention of hybrid electrode with a very high specific capacitance

of 130 F.g-1. Nano-structured BDD microelectrodes (BDD nano-grass) with a high aspect

ratio and a large surface area, is an answer to TiN and IrO2 microelectrode arrays.

4.2 Future scope

The results of the studies presented in this thesis further suggest some area of scientific and

technological interest. The in-situ activation process described in the thesis can find

application in several analytical processes such as detection of neurotransmitters like

dopamine, catecholamine, etc. In-vivo analysis is also possible as BDD is known to be bio-

inert. Other applications include in-situ activation while quantifying the total polyphenol

content during wine fermentation, waste water treatment, etc. The BDD strip electrodes and

microelectrode arrays can be used for innumerous biomedical applications such as detection

of heavy metals, neuro-chemical and –electrical recording especially to study the neural

activity, drug delivery, neural prosthesis such as retinal, cochlear implants, etc.

Functionalization of microelectrodes with enzymes can further extend the advantages of the

microelectrodes as glucose, alcohol, DNA sensors etc.

Conclusions and future perspectives

Chapter VI Page 131

Conclusions and future perspectives

Chapter VI Page 132

133

Appendix A A.1 Second order curve fitting results for UA concentration (see equation 4.1)

iA = iA0 + aCUA + bCAA + cCUA² + dCAA² + f CUACAA

Value Standard Error iA0 4.24028 1.70537

a 0.0671 0.00236 b 0.06186 0.01021 c -2.91131E-6 8.58669E-7 d -2.35667E-5 1.81746E-5 f -1.22257E-5 3.07207E-6 Number of points 18 Degrees of Freedom 12 Reduced Chi-Sqr 5.16118 Residual Sum of Squares 61.93418 Adj. R-Square 0.99798

Fit Status Succeeded(100)

A.2 Second order curve fitting results for UA concentration (see equation 4.2)

iB = iB0 + aCUA + bCAA + cCUA² + dCAA² + f CUACAA

Value Standard Error IB0 8.21117 1.21022

a 0.017 0.00168 b -0.00874 0.00724 c -1.94345E-6 6.09359E-7 d -2.66667E-7 1.28977E-5 f -8.52571E-6 2.18011E-6 Number of points 18 Degrees of Freedom 12 Reduced Chi-Sqr 2.59923 Residual Sum of Squares 31.19072 Adj. R-Square 0.97399

Fit Status Succeeded(100)

A.3 Second order curve fitting results for UA concentration (see equation 4.3)

134

IA = iA0 + aCUA + bCAA + cCUA² + dCAA² + f CUACAA

Value Standard Error IA0 53.47723 15.68371

a 0.05391 0.01588 b 0.02718 0.00552 c -2.56071E-6 5.46209E-6 d -6.69643E-9 5.089E-7 f -5.75357E-6 1.41031E-6 Number of points 30 Degrees of Freedom 24 Reduced Chi-Sqr 348.06835 Residual Sum of Squares 8353.64036 Adj. R-Square 0.91382

Fit Status Succeeded(100)

A.4 Second order curve fitting results for UA concentration (see equation 4.4)

IC = iC0 + aCUA + bCAA + cCUA² + dCAA² + f CUACAA

Value Standard Error IC0 38.40045 7.59239

a 0.01469 0.00769 b 0.02105 0.00267 c 3.28929E-6 2.64417E-6 d 5.11161E-7 2.46356E-7 f 3.92786E-6 6.82721E-7 Number of points 30 Degrees of Freedom 24 Reduced Chi-Sqr 81.56888 Residual Sum of Squares 1957.65321 Adj. R-Square 0.99144

Fit Status Succeeded(100)

135

136